Castle-Berg 50 MHz conversion

[JimCob]

Anybody else tired of waiting for 50 MHz Castle-Berg receivers? If so, then read on, because I’m going to describe my experiences with a successful conversion of the 4-channel MS4L receiver to the 50 MHz ham band.

Warning! You’re no doubt voiding your warranty by doing this and you’re running the risk of a malfunction. It’s your responsibility to ensure all of your R/C components are working properly and are flight worthy according to the AMA guidelines. I’m posting this only for the purpose of passing along information to those who also want to use the 50 MHz band for their small electrics.

For starters, if you haven't tried working with SMD components before then you might find this thread interesting as it contains lots of useful information.

SMD Soldering Thread (http://www.rcgroups.com/forums/showthread.php?t=585544)

Background
The Berg Rx uses a TA31136 FM IF chip. Take a few minutes to find a copy on the web, print it and read it. You’ll find that the local oscillator (LO) section (page 4) can be configured to use a 3rd overtone crystal, which it turns out, is what the circuit that the Berg uses. After a few minutes with a magnifier and an ohm meter I determined that the Berg circuit looks like the following.



(more to follow)

[JimCob]

Along with the crystal, the parallel L-C portion (C5 and L1) in the diagram, are the only parts that need to be changed for this conversion. You can still get the original Berg crystals (originally meant for the Berg-6 DSP III receivers) at www.rc-direct.com. These crystals are larger than the crystals used in the MS4L but they can easily be made to fit.

The Theory
My understanding is that in a Colpitts oscillator using a 3rd overtone crystal, you want the parallel L-C to be resonant in between the 2nd harmonic and the 3rd (the one you want). It turns out that my choice of components resonate well below that frequency but it also works perfectly, so who’s going to argue? Those were the components I chose as my starting point, maybe on the next one I’ll stick closer to the theory and see what happens. In other words, it worked on the first try :)

[JimCob]

I changed the green colored components (see diagram) to 330 nH and 100 pF (from mouser.com). Cut the crystal socket off of its leads leaving them as long as possible. Solder the new crystal directly to the remaining leads (see the picture). Unfortunately, while removing C5 from the board I also removed C4. Oops! When I powered up the Rx with C4 removed (but C3 in place) the LO was about 2KHz low which is not a good thing. Two options - either replace C4 with a bigger cap or short out C3. I tried the latter, and the frequency came up to within 100 Hz of perfect. Done!

[JimCob]

Flight testing
Initially I was wondering if the receiver’s front-end circuitry would cause a loss of sensitivity at 50 MHz compared to 72, but it doesn’t seem to be the case or at least isn’t significant. I tested it with my spectrum analyzer and sweep generator and it came out _better_ at 50 MHz but I didn’t have a valid test scenario because of the AGC circuit in the receiver and probably a big impedance mismatch. Maybe the AGC as it is doesn’t work well at 50 MHz?

There are probably lots of ways to improve my modification, but I don’t really want to pull parts of the board, measure them, and then try to put them back because chances are I’m going to break something in the process.

The usual antenna-down range check went at least 200 feet before I got tired of walking and turned around. No problems in the air either, I flew my test plane out quite a ways and then collapsed the Tx antenna down to about 2 sections – no problems. This morning I flew it in a small pylon racer - also no problems.

I’ll post more results once I get more flights in this weekend (assuming it doesn't rain!)

[vintage1]

Nice one mate!

Now how to convert all those hitech feathers to 2.Ghz ;)

[JimCob]

Thanks! Not too long ago I also converted a GWS 4 channel Rx. Funny thing was it still glitched! What was I expecting? :rolleyes:
Just goes to show you who has the better hardware and software design...

[dalbert02]

Jim,
That is great! I can't wait to try it. I only know enough to be dangerous... :(

Maybe you could also write an article on how to tune your tx and how to tune the rx when the antenna is lengthend. That would be much appreciated!

Also, what spectrum analyser do you use? Can you recommend one for poor people?

Thanks,
-dave

[Phil Cole]

I don't see a whole lot of front end filtering or impedance matching (i.e. a transformer) in there. Is that the case?

Since it's single conversion the image won't be attenuated much by the front end anyway, so I guess they the didn't even bother trying. Swamping by strong signals is likely to be in-band as well, so a front end filter wouldn't help that either.

[vintage1]

The only real function of an input (per mixer) filter is to eliminate out of band stuff - which can be important if te mixer responds to harmonic frequencies..and to cut down the superhet image frequency at +(or -) 455Khz..

Lots of cheaper Rx's simply don't bother with them. Slightly better ones have a single coil very loosely coupled to the aerial..those can be peaked up easily by using a retracted TX antenna and moving away and setting to minimum chatter on the servos.

Ive seen other stuff with TRAPS - LC series filters - that are designed to suck out unwanted out of band stuff..those are generally not tuneable though..and don't need to be either.

[JimCob]

Interestingly the Castle design (and maybe the older Bergs as well?) has an automatic gain control circuit driven by the TA31136 RSSI line. I'm sure this really helps cut down on the interference and distortion when operating in a strong signal environment like an indoor event.

Later on tonight I'll post the layout of the RF front end with the AGC. I don't know what all of the components are without pulling them off the board and testing them, but some good guesses can be made based on the layout. Seems like there are some LC traps in there and the rest of it may be (or should be?) impedance matching since the TA31136 input impedance is much greater than 50 ohms according to the data sheet.

Jim

[JimCob]

Jim,
That is great! I can't wait to try it. I only know enough to be dangerous... :(

Maybe you could also write an article on how to tune your tx and how to tune the rx when the antenna is lengthend. That would be much appreciated!

Also, what spectrum analyser do you use? Can you recommend one for poor people?

Thanks,
-dave

Dave - Your Rx antenna should only be a 1/4 wavelenth at most otherwise your impedance won't be in the 50 ohm ballpark.

See my post #8 in the thread below for information on how to make a shortened Rx antenna. Requires test equipment to get it right on the money.

http://www.rcgroups.com/forums/showthread.php?t=363897

Regarding the spectrum analyzer, for years I used an AIL Tech 727 giant boatanchor, but now have a $cheap chinese unit made by ATTEN. Search on ebay for them and you'll see several different kinds. I have the 6011 synthesized version, the cheaper ones look like they are analog but I could be wrong (meaning frequency drift). The 6011 works pretty well but isn't what anyone would consider a "serious" piece of equipment. If you want a better caliber of equipment then look for an older HP or Tektronics unit. For serious close-in testing you need resolution bandwidths down to 3 KHz or less which the less expensive units (including the ATTEN) won't do.

[dalbert02]

Thank you kindly! I already have the MFJ-269 mentioned in that thread, I just really need (ok want) a spectrum analyser. (sigh)...
-dave

PS Whaa! Sadly, this link referred to no longer works...
http://eweb.chemeng.ed.ac.uk/jack/radio/software/loading.html

[JimCob]

Thank you kindly! I already have the MFJ-269 mentioned in that thread, I just really need (ok want) a spectrum analyser. (sigh)...
-dave

PS Whaa! Sadly, this link referred to no longer works...
http://eweb.chemeng.ed.ac.uk/jack/radio/software/loading.html

Dave - Try it again, I clicked on the link you posted just now and it came up for me.
Jim

[dalbert02]

Thanks.
It works for me now too...
-dave

[mtucker]

Anybody else tired of waiting for 50 MHz Castle-Berg receivers? That is for sure!

I bought a couple Berg 6 on 50MHz before he sold to Castle, but it sure is nice to know about the conversion.

When did FMA stop making recievers on 50MHz? I just went to their web page and don't see any more. I have an M5 on 50MHz, but liked the Bergs better... mostly because he didn't sell the 50MHz at a premium price.

Matt

[JimCob]

Matt - I found an easier way to make the conversion, just add one capacitor and a new crystal. See this thread here: http://www.rcgroups.com/forums/showthread.php?t=676885

-Jim

[mtucker]

Excellent. It doesn't get much easier than that.
Thanks,
Matt

[barefootbass]

I converted a Berg using the add a 68pF method in the other thread. It works but the range is terrible. I was getting a range check of about 20 feet and flying range of about 200. I played around with it and changing the sync pulse width on my tx from 200ms to 210ms gave me about an 80 range check and I haven't tried flying yet. I am concerned. On my other receivers I can walk almost 2 blocks away with parked cars between me and the receiver before I loose signal. The spec sheet on the TA31136 mentions changing r51 on the LO circuit to increase the sensitivity. Any thoughts on how to do this? Or should I try the original mod mentioned in this thread first?

[JimCob]

I think the first thing to do is to see if the LO is on frequency. For example if your crystal is channel 01 at 50.820 then your LO would be -455 KHz at 50.365 MHz.
What kind of receivers and/or test equipment do you have?

[barefootbass]

My other receivers are Futaba 168f's. I have a scope and that's about it. My xtal is on channel 2. My tx is a Microstar 2000 that has no trouble on the other rx's. I have not yet tried changing the shift output.

[DPlumpe]

Barefoot-

Any progress?
Anybody been successful with Jim's conversion?
Inquiring minds want to know. Inquiring minds want receivers on 6 meters.

-Dave
K4WZA, AMA 8711

[JimCob]

If anyone is having problems with this mod I'll be glad to help out any way I can. I have 3 that are working great. (Unfortunately my Sokol died in a mid-air this past Saturday :rolleyes: but you can bet that Rx is going into my next airframe!)
Heck, if it doesn't work, send it to me and I'll debug it for you!
All in the name of flying on 6m :D

-Jim N6PA

[DPlumpe]

Thanks for the offer, Jim.

My first task is to see if I can get xtals for 50.92mHz, channel 6. RC-Direct is out of them and suggests a backorder. I've emailed to see if they think they can actually get the xtals. If they think they can, I'll get a Berg 4L and give your mod a try.

Wish I knew of another xtal source. Many decades ago I used to order special xtals from International & Texas, but I doubt that'd be affordable now.

-Dave

[JimCob]

Dave - I have some on 06, I'll trade you for a crystal on 01 (which is what I typically fly on).
Jim

[DPlumpe]

Jim- I sent you an email.
-Dave

[Goinav8n]

Jim

I have a 7 channel for you if no one has attepted the conversion on it. I won it in a raffel so Im game if you want to give it a go.

I am also looking for more 50 mhz receivers. I just ran across this thread and will give the 4 channel a shot. My DLG could really use one

Jeff
AE1O

[JimCob]

Jeff - I sent you a PM.
Jim

[JimCob]

Thanks to Goinav8n, I got my hands on a Berg-7 with end pins. The 1-capacitor mod worked so well I almost didn't want to send it back ;)
I made two test flights at a fairly busy field with my electric Extra 300. There were other pilots flying on the ham band and on 72 MHz all at the same time and I didn't have any problems with the range check or in flight.

The RF design looks the same as the 4 channel but I didn't really study it that closely. I just added a 68 pF capacitor from the 33 nH inductor to ground in the oscillator circuit. Then I soldered the crystal on top of the existing crystal socket and that was it. This one would be super easy to convert back to 72 MHz if you really wanted to. Since the crystal is larger the case won't fit back on without some minor trimming, but I didn't bother with that I just wrapped the Rx in foam.

I'll upload a picture as soon as I can get it scaled down to a reasonable size.

[Goinav8n]

I flew my modded 50 mhz Berg in my 3D Katana today in front of the house. Solid with no glitching. Works good so far. Ill be abusing it in this and a few other planes soon

Jeff AE1O

[jpeca]

These simple conversion sound very interesting. Has anyone tried converting a GWS 4 or 6 RX? Iwould like to try it on some older equipment first. Also, does TX brand (negative VS positive shift) make any difference?
Jeff

[JimCob]

Jeff - Yes it will work, in fact my first conversion was a GWS 4-channel. Same technique applies though the component locations differ and you might need to replace the inductor as well (I think I did). Funny thing was that my GWS Rx used to glitch every now and then on 72 and the move to 50 MHz didn't help any!

-Jim

[gigelus2k3]

Why would one use a 50MHz RX in US? Is it even legal to use it for remote control?

B.t.w., I recently fell in love with SombraLabs' Lepton line of receivers. They also sell a 50MHz variety and I'm assuming is of the same high quality as their 72MHz flavor.

Serban

[JimCob]

Its legal as long as you have a Ham Radio license. One of the benefits is that there are usually fewer people on those frequencies so less waiting for a frequency pin at the field.
I agree with you about Sombra Labs, they make some nice receivers.

-Jim

[gigelus2k3]

How about the TX, are you retuning it too for 50MHz? I guess that there are no "channels" defined for that HAM band, so every one could pick the frequency he wants.

Serban

[JimCob]

Actually there are recognized channels just like on 72 MHz. Type "radio control frequencies" into Google and you'll find lots of lists. Both Futaba and JR make gear for the Ham Band. Personally I fly a Futaba 9CAP transmitter with a Futuaba module.

[R.A.P.]

Barefoot-

Any progress?
Anybody been successful with Jim's conversion?
Inquiring minds want to know. Inquiring minds want receivers on 6 meters.

-Dave
K4WZA, AMA 8711

Jim/and Dave
I just today got a chance to convert a Berg7P per Jim's instructions, and
I have to admit I struggled with the soldering of the 68pf, took me a good
30 minutes!!! the next one will be easier.. thanks Jim for this informative
thread.......
Dick WB6NOG ( I know I'm a little slow in my responce)

[R.A.P.]

I just finished a Berg4 microstamp rcvr. and it took about 5 minutes for
the 68pf and crystal mod. and it works great... thanks again Jim.
Dick

[JimCob]

Dick - Glad to hear it worked for you! By the way, what kind of transmitter are you using?
Jim

[R.A.P.]

Hi again--- I'm using the Futaba T14mzAP, I use this xmtr. on 72MHz,50MHz and 2.4GHz,the only fault I can find is that I some times feel more like a programmer than a modeler!! We need to find a crystal mfg. to buy the
smaller 50MHz crystals, I have been giving some thought to calling Castle
and asking them where they buy their's.
So blue sky's and light winds.....
Dick

[JimCob]

There is! Sombra Labs (www.sombralabs.com) sells the smaller sized crystals for 50 MHz and 53 MHz and I've already done some integration testing on the 4L.

There are 2 capacitors involved, the 68 pF as before plus another one to bring it down a bit to the correct frequency. I've only converted one so far for another ham who is currently flying it. I'll check with him about how well it is working and then post it here.

-Jim

[R.A.P.]

Jim-- sounds good on the crystals....In the mean time I sure need to
sharpen up my soldering ........
Dick

[R.A.P.]

I flew the Berg 7 yesterday at Sea World field and I could not be more
pleased., the range was as for as I dared ( plane VERY small to see) and
no glitches or problems. I can't figure out why Castle did'nt go ahead with
the receiver? however I have heard that they were coming out with a 6 meter
receiver of some sort in the future!!!!!!
Dick

[JimCob]

Dick - That's great! I'm flying several and really don't even think about reliability anymore, they just keep working. Strangely I haven't heard back from the other guy yet regarding the Sombra crystal conversion I did for him.
-Jim

[JimCob]

Dick - One more thing, there is a thread in the Castle forum regarding the 6m receivers and the latest posts say that they're not going to do it, not enough volume to make it worthwhile I guess.
-Jim

[wb6hro]

Jim,

I 've ordered a Berg 7P to do the converstion. I am going to see how close the LO is using one of my FMA M5 xtals. I have an IFR 1200S service monitor and will look at it. Would be nice to know what the loading on xtal is for the correct freq.

If the FMA is close or I have to get another one from RC direct, I think plan on taking it Cal Crystal Labs in Anheim to get some more made and they can test the xtal so we know what the loading is. I have to get some spare crystals made for my 220 repeater which uses a VHF High Band GE MVP that I converted to 220 anyway.

De Joe Orrico WB6HRO

[JimCob]

Joe - I don't think the FMA M5 crystals are going to work because they are dual-conversion types and the Castle Rxs are single conversion. They crystals that I know to work are the old Berg types from Rc-Direct or those made by Sombra Labs (www.sombralabs.com) that are in the smaller UM-5 case.
I did a Berg-7 conversion for another ham using the Sombra crystal and had to change a second cap to get it closer to the right frequency, probably because the crystal specs that Sombra Labs used differs enough from the Berg design to matter. I didn't fly it using the orignal modification (before adding the second cap) it just seemed to be off far enough on the bench that I was motiviated to try to fix it.

[mtucker]

They crystals that I know to work are the old Berg types from Rc-Direct or those made by Sombra Labs ..... RC Direct still has most of the 50MHz crystals in stock. The one they don't have is of course the frequency I fly on CH7 (50.940MHz). I recently (4 months ago) bought two Berg 6 from Mr. Berg (he lives a few miles away from me) and he didn't have CH7 in his private stash. After I told him RC Direct was out, he told me to by the JR crystal (gray case = single conversion). I put them in the other night and all worked well.

The JRs are more expensive than RC Direct and you have to pull them out of their plastic case. http://www.horizonhobby.com/Products/Default.aspx?ProdID=JRPXFR07

Matt
N6EAJ

[wb6hro]

OK got my new Berg Rx yesterday and picked up some 68 pf chip capacitors.

Fired up the service monitor and yes Jim you are correct the FMA is dual conversion while the Berg is single conversion. So I was ready to order a crystal from RC direct and found an old Futaba PCM - 512 rx I had and opened it up and gee it was single conversion on channel 06 50.920! You never get that lucky!

So I pulled the xtal and tacked soldered it in and low and behold it was within 30 hz of the correct mixer freq! Anyway it was labeled "Type 1" on the Futaba xtal. Anyway I think I will take it over to Cal Crystal labs and have them make me a couple in the smaller case and we will have the loading so anyone can order what they need in the future.

I also test both my tx modules on 04 and 06 to see if they were on freq. Both were about 1 khz low in freq which is alot when you are only talking about 2.2 khz of deviation of the FM signal. Will have to look and see if there is a means to tweek them back on freq.

Anyone need a 6 Channel Futaba PCM-512 Rx? It is a FP-R116GP on 6 meters. I don't have a 512 xmiter anymore, free for the shipping.

[JimCob]

Great news, I'm glad that worked for you. It'll be nice to know the crystal parameters also as they might come in handy for future reference. Let us know what it costs you to get your new crystals made.
Regarding your Tx, Futaba transmitters use a negative shift, and due to the nature of the modulation it spends a little more time at the lower end than at the other- in other words it isn't symmetrical and reads low on a frequency counter. 1KHz low is probaby about right depending on how many channels you have.
For tuning I made a little box that connects to the trainer port which allows me to toggle between the high and low frequencies so I can adjust each one separately. The Futaba Tx modules have pots for that purpose.

-Jim

[Cincinnati Joe]

Got my Berg in the mail today. Took the grey plastic case off my JR Ham 05 crystal, bent the leads and soldered them directly to the board where the old crystal socket went (I unsoldered it carefully).

Powered up and nothing worked. I only had "on hand" junk parts to work with, so first tried a 50pf disc and it sputtered a little and the servo sort of followed the stick... so I figured if 50 wasn't enough, bigger was better. The closest thing I had in my junk bin was 150 pf, so tried that, and damned if it didn't work!

I'm using this with a JR10X transmitter. With the antenna collapsed to just one section, I stopped walking at about 150 feet. This is with reciever sitting on a cardboard box about 12 inches off my garage floor, next to a metal door!

I figure I should easily quadruple the range with the antenna up and the plane in the air, and that puts it well beyond the range of any indoor venue!!

I only wish my 10X would hold more than 10 Planes!~ now I've got to get the "datasafe" to work with VISTA - which might be a lot harder than converting the receiver!!