DIY Lipo charge guard

[hul]

just finished building and testing this Lipo charge guard:
http://mujweb.cz/www/coro/Detaily.htm
originally mentioned here:
http://www.rcgroups.com/forums/showthread.php?t=388213

It uses a FET in series with the pack, any cell over 4.2V starts to turn the FET off (increases its resistance), which fools the charger into thinking the pack is full. The charger then stops the charge as per normal.

All things normal, charger set correctly and the pack only slightly out of balance, the FET only needs to add a few 10mV drop. It therefore only heats up very little.

I charged a couple of full 3s packs at 4s to really test it.
The 2200mAh pack stopped charging with "wrong cell count" alarm, the FET was not really hot after this. The 4400mAh pack heated the FET enough for me to stop the test. The FET needs overtemperature protection to be foolproof.
In both cases none of the cells exceeded 4.20V.

The design scales up to any number of cells without having to use different components. Mine is for 6s and fits easily in a 6 x 4 x 1.5 inch case.

Hans

[AndyOne]

Hans,

It sounds like the FET is not being turned on completely if it heats up. The FET should act like a switch being fully on or fully off. When it needs to be on it is important to make sure there is enough voltage to the gate to ensure it is fully on otherwise it will act like a resistor (>10mOhm).

It may help if you were to show us the circuit of this device, I followed the links but on a quick look I didn't find it, my patience isn't great when following links if I can't find in a couple of clicks I usually give up.

Andy.

[GlowFly]

The FET really needs to be on a modest heatsink for this circuit because as a cell approaches 4.2v the FET enters a linear mode of operation. The actual result depends on the charger characteristics.

Another possible cause of heating is due to instability. A poor layout of the FET gate lead can cause self oscillation. Ideally the gate lead should have a small series resistor, say 1k, close to the FET if the leads are a few inches long or more.
--
Steve

[hul]

the FET is meant to be in linear mode, to drop the excess voltage the charger puts out. Diagram borrowed from http://mujweb.cz/www/coro/Detaily.htm

The circuit works perfectly. I think, I didn't make myself quite clear....
Heat is expected when charging 3s packs at 4s...

I don't think it needs a heatsink. A TO-220 has about 60degC/W thermal resistance junction/ambient. My largest pack charges at 4.5A. It would take 222mV total cell imbalance to heat it up by 60degC. I'll take action before 222mV cell imbalance is reached.

It would need a massive heatsink to cope with incorrect charger setting or dead cells; cell count set to high by one would result in about 19W for my largest pack. Wouldn't need such a heat sink if it had overtemperature protection.

Hans

[hul]

changed the FET to a VNP35N07 with built in over-temperature protection. Had to change the 47k resistor to 23.5k (47k//47k) because the gate draws a little current in this new FET (to supply the internal supervision circuit).

I think the charge guard is pretty much fool proof now.

Hans

[coro]

I think the charge guard is pretty much fool proof now.
Hans
Dear Hul,
Great idea to use temerature protected fet! Yes, I recommend not to use the heatsink, I had just thermal fuse over the classic fet yet.
Problem with keeping too high voltage too long and overheat the fet (that I saw with Robbe chargers) is solved by temp protected fet, so for all others I am willing to add this Your idea onto my www soon.
I hope that You will like unit's real behaviour. There are side effects with such simple module, like lowering voltage by warming activated TL431 by cca 2mV, which helps with reballancing pack.. I hope all your pack will keep ballance even with so low corrective current.

There is one more important thing to say:
Some chargers can rise their voltage over safe limits and destroy themselves or used fet transistor by overvoltage. I am using another power transistor connected parallel to input of the module, (with resisor or small halogen bulb on + side), with base connected thru led (1,5V voltage difference alowed, led marked CHARGER ERROR ) and protective resistance to module - output.
This will prevent input voltage from rising too far over output (=pack) voltage, and thus keeping heat produced much lower. This is preventing high voltage damage possibility to both charger and the limiter.
Yes, this is also the sollution to problems I saw described with polyquest PCM guard and Triton chargers.
Best regards.. Coro

[hul]

the unit works really nicely. This is what happens (in about the time it takes to read this):
a cell reaches 4.2V, LED lights up dimly, FET increases resistance, charger output voltage increases to 12.7V (from 12.6, 3s setting), charge current reduces slightly, charger output goes back to 12.6V. No sign of any instability.

At no time was I able to get more than 4.20V cell voltage, even when selecting the wrong cell count and high charge current. I think that's quite an achievement.

My Schulze 636 puts out a measured 65V after disconnecting the pack mid-charge. The FET sees that minus pack voltage, about 53V for 3s, which is safe for the VNP35N07. Other chargers may be different.
The charger may not be fit for service, if it gets destroyed by this.

The temperature protected FET was Acetronic's idea, not mine.
The drawback: it draws gate current. The charger detects that as a connected battery and then puts out alarms (wrong cell number selected etc). Will have to connect things in the right order, pack to charge guard, then charge guard to charger.

Thanks, Hans

[coro]

The drawback: it draws gate current. The charger detects that as a connected battery and then puts out alarms (wrong cell number selected etc). Will have to connect things in the right order, pack to charge guard, then charge guard to charger.

Thanks, Hans
Warning about short circuit when two male 4mm jacks can touch before connecting to charger! there is full pack power on them..
Maybe external temperature fuse can be less problematic :( ?

[hul]

Warning about short circuit when two male 4mm jacks can touch before connecting to charger! there is full pack power on them..
you're two days late. Let the smoke out of my first FET that way....

Would you have a part number for that external temperature fuse? It would solve the issue, but would it be fast enough?

Hans

[hul]

Warning about short circuit when two male 4mm jacks can touch before connecting to charger! there is full pack power on them..
you're two days late. Let the smoke out of my first FET that way....

Would you have a part number for that external temperature fuse? It would solve the issue, but would it be fast enough?

Hans

[Acetronics]

Hi, Hul

Why not use the Raychem Polyswitches ??? it switches off for over-current AND also, for a 120°C temp.
glass devices nested into transformers or Electric motor windings ( GTE from Sylvania ) do the same ...

just "stick" one on your Mosfet ...

Alain

[coro]

Hi, Hul

Why not use the Raychem Polyswitches ??? it switches off for over-current AND also, for a 120°C temp.
glass devices nested into transformers or Electric motor windings ( GTE from Sylvania ) do the same ...

just "stick" one on your Mosfet ...

Alain
I have no part number yet, I found and buy some in local store just for testing. Those are used also in batery packs.
I am using it to disconnect fet's Gate from + voltage, because I do not want any resistance to be added into charge circuit. Fet has 0.06Ohm or less..
Raychen can be probably used the same way. Low power can be used to temperature protect by interrupting Gate power. High current fuse (10A should be adquate value) with very low resistance can probably be inserted series to fet, it depadns on its resistance which i did not found..?
Over-current protection is problem in case You have to work without any voltage drop. Everything can be solved, (using induction, hal sensor or so..) but this unit has to be simple and small and near-zero voltage drop is important...
Every Fet is equiped with reverse protective diode, so closing the fet is not usefull to help with short circuit protection with connected pack.

I have this unit build into charge wires, and I am not disconnecting it from charger. You can charge and also discharge thru this module (its fet in the path) with currents cca 5A (depend on used fet), and it makes no problem when You charge/discharge NiCD pack (without connection of voltage monitoring).
If You use same type connectors on input of the module (i call it "Limiter"), You should care them the same way as original pack power connectors, and use original charge wires from charger. Short circuit is not acceptable with Lipo pack, and destroyed fet is much better than destroyed pack.

[hul]

I am using it to disconnect fet's Gate from + voltage, because I do not want any resistance to be added into charge circuit. Fet has 0.06Ohm or less..
Raychem can be probably used the same way. Low power can be used to temperature protect by interrupting Gate power.
Haven't tried this, but I think the Polyswitch won't have enough resistance when tripped to turn off the FET. Or you need a small (low resistance) resistor between gate and source to pull gate voltage down. In that case there is current flowing again, which makes the charger think there is a battery.

Alain: I'd prefer to turn the FET off instead of having an inline fuse.

Hans

[Acetronics]

Hi, Hul

I saw there were fets with temp and current info ...IRC 540 i.e
But, you'll need extra circuitry !!!

End of thread for me ... bye bye

Alain

[coro]

Haven't tried this, but I think the Polyswitch won't have enough resistance when tripped to turn off the FET. Or you need a small (low resistance) resistor between gate and source to pull gate voltage down. In that case there is current flowing again, which makes the charger think there is a battery.

Alain: I'd prefer to turn the FET off instead of having an inline fuse.

Hans
Dear Hans,
Unit resistance that can discharge pack is not nice and I do not like it. Of course I am not willing to let this unit connected to pack for a long time, but in fact there is about 0,5mA discharge which I do not want make higher.
As I said before, without inline fuse unit cannot be easily protected against high discharge current when charger connection is shorted, but it takes just the same care as the original pack cables.

There is really working thermal sollution with old good bimetalic relay type temperature fuse, I am using SELCO product now. I still can not find useable partnumbers and descriptions, and I like idea with VNP35N07 much more.

I am not sure about Your Schulze behaviour. It will not accept connected resistance (i think that ZERO voltage is there) on its output before the pack will connect? I look at the VNP35N07 datasheet, and it should draw typicaly only 250 mikroA.. I am using Zener diodes to protect IRLZ fet's gate, so some low current flows thru my original module too... :(

[coro]

Hi, Hul

I saw there were fets with temp and current info ...IRC 540 i.e
But, you'll need extra circuitry !!!

End of thread for me ... bye bye

Alain
Dear Alain,
Thank You very much for Your valuable idea.
Btw, using two such temperature and current protected fets connected serial in opposite direction (sorce to source) and connected gates, could be the easiest way to make short circuit protection? Its just idea, but I thing I should try something.
Coro

[hul]

I am not sure about Your Schulze behaviour. It will not accept connected resistance (i think that ZERO voltage is there) on its output before the pack will connect? I look at the VNP35N07 datasheet, and it should draw typicaly only 250 mikroA.. I am using Zener diodes to protect IRLZ fet's gate, so some low current flows thru my original module too... :(
the charger puts out 10.2VDC and 0.9VAC on top of that with nothing connected. There was no problem with your original circuit using a 12V zener diode.
I found a bimetallic normally open temperature switch in a TO-220 case, Farnell part number 279808 (http://au.farnell.com/jsp/endecaSearch/partDetail.jsp?SKU=279808&N=401). Not cheap, but it will make the circuit bullet proof. Will bolt that back to back to a conventional FET, wire it between gate and source and leave the rest of the circuit unchanged.

Hans

[coro]

the charger puts out 10.2VDC and 0.9VAC on top of that with nothing connected. There was no problem with your original circuit using a 12V zener diode.
- This looks reasonable for maximum of 4 LiPo cells. But it is probably just detection of anything connected to start another charge. Will anything happen if You will connect a pack with exact 10.2V rest voltage?

I found a bimetallic normally open temperature switch in a TO-220 case, Farnell part number 279808 (http://au.farnell.com/jsp/endecaSearch/partDetail.jsp?SKU=279808&N=401). Not cheap, but it will make the circuit bullet proof. Will bolt that back to back to a conventional FET, wire it between gate and source and leave the rest of the circuit unchanged.
This is really simple and sure. I call such things an "elegant sollutions".

I believe that advanced risc procesor with object oriented programming is not allways the best sollution to limit simple voltage.

Have a nice day!

[hul]

Will anything happen if You will connect a pack with exact 10.2V rest voltage?
I think it's the AC component that detects the pack. It's probably high impedance and drops close to zero as soon as something is connected.

Hans

[coro]

Alhtrough teoretic phase of project done, the real testing still continues by me.

In the real life the CHARGER behaviour wary by kind of charger.
Some of good chargers work perfectly. I have tested Orbit, Simprop, Jamara and Shultze without any problems.

But som intelligent chargers will try to measure voltage without current.

This idea is not my favorite, I do believe that slow damage of cell can occur:
"An overcharge of 0.1V can cause the excess lithium to move from the cathode to the anode by a process of electroplating."
http://www.rcgroups.com/forums/showpost.php?p=4368403&postcount=20

At 1st , chargers are not measuring voltage constantly.
- depending on frequency of voltage measurement, I have tested the charger that was putting about 37-40V onto this unit, no matter that setting was 2s LiPo. Pack is protected by Limiter unit well enough, but huge overvoltage means huge heat produced.

At 2nd, some chargers will rise current after it was dropped , again and again.
- I have never heard about rising current on any LiPo charge graph, this is very bad charge algorythm...
- as tested for example on Duratrax ICE, this results in temporary overvoltage on charger output. (ICE reduces current almost immediately, so ICE can be marked as quite Good charger, too.)

All in all , excess current provided by charger is such occasions, is relatively small. Most of the current is still accepted by pack (those which passes thru Limiter) .
Amount of current not alowed to pass thru Limiter can be easily consumed by one single npn transistor paralely connected to charger output. Base of this npn is connected to Limiter device negative output thru some resistance about 1k, up to 2k2 works well. (protective resistance on + side of about 1-5 Ohm should be used also)
Now - voltage on charger output can rise only by 1-2V, so heat produced is acceptable ( only few mA current, but the overall voltage is applied ).

Personaly, I did thow away the charger that I was found to do this, but since it is Robbe, I have to try to adapt device to any situation that is possible.
Limiter without this addition willl stop the charge due overheat, just a little mAh sooner than at 100percent full cells.

This post is meaned as a warning to all the builders of Limiter - Your charger quality matter.

[coro]

Just tested Raychen Polyswitch 020 , 200mA 100st Celsia version. When temperature is below 100st, its resistance is low (4..6 Ohm) as shown in the product sheet, and at overtemperature its resitance rises over my DVM range. This resistance is high enough but I can not find its value in datasheets.
When inserted into + side of the power part of Limiter , and second resistor of up to 220 kOhm is added to discharge gate of the fet. In case of overheat the unit will disconnect charger completely, at least until temperature drop. Charger should never start again without user attention.
Polyswitches are small, genarely available and much cheaper than fet with integrated protection. You should use smallest size available, as we are talking about driving the fet's gate.

[hul]

I considered doing that but was worried that the Schulze would detect the 220k resistor as a connected pack (like it did with the Temp FET) and start charging. It's a very small current, it may work.

Hans

[doberman]

Just started reading (finished) both of the threads on this limiter/balancer - since the last addition to this thread (end of Sept) wonder if more testing has been done or any other componets have been used? Would a limiter /balancer like this work with the Graupner charger that sells for $25 - if I remember correctly the Graupner is a PULSE type charger - what chargers would you reccomend or not reccomend this device to be used with?

[hul]

Just started reading (finished) both of the threads on this limiter/balancer - since the last addition to this thread (end of Sept) wonder if more testing has been done or any other componets have been used? Would a limiter /balancer like this work with the Graupner charger that sells for $25 - if I remember correctly the Graupner is a PULSE type charger - what chargers would you reccomend or not reccomend this device to be used with?
I have this device in regular use since I built it, charge about 5 packs per week on average. All of those packs have remained in balance (20mV or better).

Did not change any components; it's Coro's original design with the addition of the temperature switch, a 100V FET (instead of the original 55V type) and some small value changes that don't affect operation.

Haven't done any test with other chargers and don't know if it would work with pulse chargers. It will limit voltage with pulse chargers, not sure if the charger would like it and/or complete the charge.
Are you sure Graupner sells pulse chargers?

Hans

[coro]

Just started reading (finished) both of the threads on this limiter/balancer - since the last addition to this thread (end of Sept) wonder if more testing has been done or any other componets have been used? Would a limiter /balancer like this work with the Graupner charger that sells for $25 - if I remember correctly the Graupner is a PULSE type charger - what chargers would you reccomend or not reccomend this device to be used with?
I did not change any component except adding thermal protection.
This design was tested before I did published it, with different components, including more and less powerfull balance currents, or with slower response to cell voltages, and published scheme behaves best of them.
After successfull tests of serial and paralel connections of 4-cell modules onto 8s or 12s, the Limiter is just in use, by those who own it.
I consider testing phase finished, just the list of chargers that cannot be used with this device should be updated.
I found problems only with Robbe PowerPeak Ultimate-Li and the same design Robbe chargers, but this is not report about mallfunction of Limiter, this is report about unacceptable charger behaviour. I know that this is attempt to workaround resistance in long thin charge wires, but unacceptable.
Hul: I am glad to hear that You are satisfied after long enough time.

Classic Graupner chargers Ultramat 10,12, 25... are good chargers with response fast enough to cooperate with Limiter.
But - the attempt to set current automatic by $20 charger as the first in the world is very bad idea.
I cannot see what current will so called "quattro" automatic charger decide to charge pack, when connected thru protective fet transistor. The current will be smaller, so charge will be slower - for sure. Voltage is limited correctly so I am not expecting troubles. As I read the thread about this charger behaviour, I see that nobody can expect charge speed even when pack is directly connected to this charger.

While I am trying to clean up the LiPo charge from unsure assumptions and probabilities, this charger is considered unacceptable by me.

[hul]

just for clarification: I only changed component values, because I brought home the wrong ones from the store. Colour blind, can't see the colour code on resistors....
There is nothing wrong with Coro's original values.

Hnas

[flieslikeabeagle]

Anyone care to clarify any differences between Han's circuit and the commercially available Polyquest charge guards?

I appreciate Han's circuit can put the FET into a linear mode as essentially a variable resistor, while the PQ device is operated digitally, essentially turning off as soon as any one cell exceeds 4.2 V. But it seems both achieve the same result - no single cell will be allowed to go over 4.2 V at any time during charging.

Yes? No?

-Flieslikeabeagle

[hul]

the PQ charge guard turns off at 4.35V+/-.025V, see http://aircraft-world.com/prod_datasheets/pcmguard.htm. Its FET is rated for 30V only, many chargers exceed that when the charging circuit is opened.
Coro's circuit (not mine, I just built it) limits cell voltage to 4.20V (or whatever value it's set at). To my knowledge the only device that does that and works with any charger. FMA's charge guard may do the same, but not sure.

Hans

[flieslikeabeagle]

Thanks, Hans! And thanks, Coro!

-Flieslikeabeagle

[coro]

Anyone care to clarify any differences between Han's circuit and the commercially available Polyquest charge guards?

I appreciate Han's circuit can put the FET into a linear mode as essentially a variable resistor, while the PQ device is operated digitally, essentially turning off as soon as any one cell exceeds 4.2 V. But it seems both achieve the same result - no single cell will be allowed to go over 4.2 V at any time during charging.

Yes? No?

-Flieslikeabeagle

CC-CV type charge I want, runs with condition MAXIMUM(cell1,cell2, cell3...) <4.20V,
instead of ordinary charger, using SUMARY(Cell1, Cell2...) < 4.20*number of cells.
Diference between those conditions is here used as voltage drop across the fet. My target is to fool the charger by increasing voltage it reads, to make it slow down.

In this case, voltage of my Limiter can be lower than voltage what charger uses. I am using 4.165V , most of my fiends requested 4.150V.
If any charge guard switches on and off to allow or stop the charge, it have to be preset to bit higher voltage than charger puts on the cells in CV phase.

[Dan Baldwin]

Anyone care to clarify any differences between Han's circuit and the commercially available Polyquest charge guards?

I appreciate Han's circuit can put the FET into a linear mode as essentially a variable resistor, while the PQ device is operated digitally, essentially turning off as soon as any one cell exceeds 4.2 V. But it seems both achieve the same result - no single cell will be allowed to go over 4.2 V at any time during charging.

Yes? No?

-Flieslikeabeagle

Beagle, another difference is that Coro's circuit will attempt to balance the pack as well as acting as a charge guard. The PQ device is strictly a charge guard.


Dan

[flieslikeabeagle]

Dan, I see nothing in the schematic that suggests this ability - the only load across any individual cell being the corresponding comparator itself, and the only effect any comparators output has on the charging process being to reduce current flow into the entire series chain. :confused:

Am I missing something?

BTW, I have been using a combination of a Polyquest charge guard and a set of lipo balancers to charge my old PQ batteries for a long time now. I started a thread in, I think, the "Batteries and chargers" forum on this a long time ago.

-Flieslikeabeagle

[Dan Baldwin]

Dan, I see nothing in the schematic that suggests this ability - the only load across any individual cell being the corresponding comparator itself, and the only effect any comparators output has on the charging process being to reduce current flow into the entire series chain. :confused:

Am I missing something?

BTW, I have been using a combination of a Polyquest charge guard and a set of lipo balancers to charge my old PQ batteries for a long time now. I started a thread in, I think, the "Batteries and chargers" forum on this a long time ago.

-Flieslikeabeagle

Yes, Mr Beagle, I believe you are missing something. If you look at the schematic of the per/cell part of the circuit, you will see that when the TL431 first starts to conduct as the voltage reaches 4.2 volts, the first thing that happens is that voltage is placed across load resistor R4. Only if the voltage continues to rise does the circuit begin linear shutdown.

Yes, I remember reading that you used balancers and a charge guard together. I don't know why that hasn't caught on more.


Dan

[coro]

I can just comment this - it is easy to add pnp transistor in place where led is connected, to achieve some higher current for balancing. LM431 (or TL431CLP used) can dispatch some amount of heat, 125mW in my case. R4 was choosen with this value in mind, and also some bias current is necesary for stable operation.
I really want this unit to be simple and small, to use it as a bit larger connector to pack wiring.

I do not want its rebalancing current to be higher for some reasons - mostly because normal pack does not require balancing, and I really want to get warning that someting is wrong (I will see lower pack voltage after charge). In addition, wiring to balancers are often poor, long and thin, so there is possibility of voltage drop in case of any higher current flow. Unit reaction is within 2-3mV now, the only bad point is precise setup of trims, full range is between 4.10V-4.30V but every TL431 has a bit diferent voltage.
Another reason why not balance is diferent IR and/or capacities of cells in pack.
Choosen values gives 20mA balancing current. It is not enough to correct serious imbalance (as it means maximum of 20mAh per hour), but it is enough to keep healthy pack well fitted or fine tune not so exact balance achieved by separate cell charge.

[flieslikeabeagle]

Dan - thanks, I had a total mental meltdown before. I didn't recognize the 431C as a precision Zener/shunt regulator, rather I thought the circuit was using several comparators, and floating the inputs of each to the level of the individual cell.

Anyway, I guess I'll go beat myself over the head with a ripe banana, maybe it will restore my mental faculties. I hope so, otherwise I may soon be stocking up on "how to open a milk carton for Dummies", "Microwaving TV dinners for Dummies", "Turning on electric lights for Dummies" and a few other similar books I might need to stay alive in my disoriented mental condition. :eek:

Geez, I feel like the confused blue fish in "Finding Nemo". (Did anyone else notice that virtually all the characters in that movie were male? And of the only two female characters, one was killed within seconds of the movie starting, and the other was essentially insane? Wonder if that's a preview of the world as Disney sees it.)

Coro - I finally got it, thanks for the clarifications. Of course, the LED indictors and the LED's in the opto couplers pull a few mA from those packs, too, to help out R4 a bit. :)

-Flieslikeabeagle

[Dan Baldwin]

Dan - thanks, I had a total mental meltdown before. I didn't recognize the 431C as a precision Zener/shunt regulator, rather I thought the circuit was using several comparators, and floating the inputs of each to the level of the individual cell.

Anyway, I guess I'll go beat myself over the head with a ripe banana, maybe it will restore my mental faculties. I hope so, otherwise I may soon be stocking up on "how to open a milk carton for Dummies", "Microwaving TV dinners for Dummies", "Turning on electric lights for Dummies" and a few other similar books I might need to stay alive in my disoriented mental condition. :eek:

Geez, I feel like the confused blue fish in "Finding Nemo". (Did anyone else notice that virtually all the characters in that movie were male? And of the only two female characters, one was killed within seconds of the movie starting, and the other was essentially insane? Wonder if that's a preview of the world as Disney sees it.)

Coro - I finally got it, thanks for the clarifications. Of course, the LED indictors and the LED's in the opto couplers pull a few mA from those packs, too, to help out R4 a bit. :)

-Flieslikeabeagle

What are friends for if not to rub each others noses in the smallest of mistakes :D .

"how to open beer for dummies"
Oops, That reminds me. I'm late for a lesson.


Dan

[coro]

Finaly, after over than year of my use and advertising of this unit, I found the charger which promises to make this design obsolette (or at least not the only single one). FMADirect Cellpro 4s.
Still not tested, however explained enough.
The only important condition, it is keeping cell voltage below 4.20V in any situation, is promised. Answer to my question whether they did sure workaround of higher resistance of balance connector was answered yes.
It is in addition cheap enough to not hesitate with purchase. All other additive functions could help with marketing and were not so important to me.
For explanation of that was achieved never before with other systems, here is simple drawing of worst case, achieved by paralel chargers (or balancer) connected by non-zero resistance wires to cells: http://mujweb.cz/www/coro/paralelpruser.gif

[coloradoz]

anyone got a good link to the schematics for this?

[hul]

seems the original link is dead.
The schematics are in this thread. Resistor values are:
R1 = 15k
R2 = 2.2k
R3 = 22k
R4 = 82ohm
R5 = 82ohm
R6 = 380ohm
R7 = 220ohm
I think I used a 10k pot, multiturn in my case

Hans

[hul]

added the missing resistor values in the above post

Hans

[coro]

Checked links and they work... however, copy is here: http://members.chello.sk/sklenar/
I did use Piher PT6 trims, 5k, simple 6mm, just because I found that multiturn ones did change values after mechanic shock..