TG1 - advice from any one who has built it...

[briomusic]

you are aware that you should have one 2N4910 and one 2N4898 each on the out amp, right? how about going to the output level pcb, are you getting signal at S3-1?

 

[wmtunate]

I had the exact same problem.  Make sure you use an isolator underneath the output transistors.  I had mine soldered down and I was getting contact with the shell and the middle pad on the top of the board.  I desoldered the transistors, cut out some small circles of rosin paper, lined them up on the board, and pushed the pins through the paper.

 

[echotodd]

Thanks guys,

I do have signal S3.  I have mica insulators underneath the transistors, the transistors are placed correctly.  I even tried disconnecting the screws I used to mount them to the PCB and I'm still showing nothing at the unbal. output.  I even tried a little sleep, that didn't help much either!  Diodes look good and the caps are oriented properly.  Time to go to work, more digging later.  Thanks again guys!

 

[echotodd]

Thanks for the tip!  It was ages ago that I stuffed the transistors but I remember checking the datasheets and everything looked okay.  I'll double check when I get home, though I would think that if the 214s weren't right I wouldn't be getting a signal at S3, which feeds the ouput amp.  I also re-checked the pin-out for the 107/177s this morning, looks good there too.  Something has got to be funky at the amplifier, I just haven't figured it out yet.  The fact that I did it twice tells me it's super obvious.  I metered for continuity between the shell of the transistors and the solder tab wmtunate mentioned in case my insulators had failed but it looks fine.

 

[crisotop]

I think (as far as I remember, built mine months ago) that there shouldn't be any insulation between the output transistors case and the mounting holes on the pcb, as this is the emitter connection.

but please double check before trying
Christoph

 

[wmtunate]

I even tried disconnecting the screws I used to mount them to the PCB and I'm still showing nothing at the unbal. output.

That may be the problem.  You may have too much isolation.  Make sure you're also getting continuity between the shell of the 4910 and T_OP right next to it.  The shells of those things need continuity to the screw pad.  The shell is considered "Pin 3".

I see crisotop already beat me to it.

 

[echotodd]

Hmm,  I'll check it out when I get home!  If I remember correctly I was showing 28V at the shell of the 4910, 4898, as well as T-OP with respect to ground.  I thought when I looked at the boards that pin 3 was on the same line as the pads around the screws to lock the shells down.  This also could be where I over thought things and it all went hopelessly wrong!

 

[echotodd]

Well thanks a ton guys!  I really hope that takes care of it as it certainly qualifies as something obvious and something I did twice (four times actually but we'll just say twice as my ego is pretty bruised about this already).  So much for erring on the side of caution!

 

[maxwall]

Have you checked all diodes for proper orientation ?

ocasionally about 1 in a 1000, I find a bad diode or one that was banded in reverse.
sometimes the body can crack very unnoticeably due to bending leads

 

[echotodd]

Thanks Frank,

I know all the diodes are properly oriented.  I would hope like heck I don't have cracked diodes across two boards!  I'm really hoping that it's just a matter of the insulators.  They seemed like such a good idea when I put them in!  I'm pretty eager to get hope tonight and tear it apart!

 

[crisotop]

prepare yourself, as this is probably my most favorite compressor ever - I was so blown away after the first use (still need to finish my frontpanel though, and haven't had time to try stereo linking)

good luck!
christoph

 

[nielsk]

Tekay, did you determine if BC500c &560C are the correct pinout?

 

[echotodd]

Always the obvious.  The issue was not insulators my dear friends.  It turns out the lock washers on the kepps nuts I was using overshot the pads.  Now my only issue is that I am slamming the hell out of my input.  My guess is that I have the hold wired wrong.  To be honest, I'm embarrassed.

 

[wmtunate]

Always the obvious.  The issue was not insulators my dear friends.  It turns out the lock washers on the kepps nuts I was using overshot the pads.  Now my only issue is that I am slamming the hell out of my input.  My guess is that I have the hold wired wrong.  To be honest, I'm embarrassed.

Been there too.  Though in my case, I tightened them down enough that they went through the soldermask and into the ground plane.  I ended up putting the nuts on the top.

 

[echotodd]

I just wanted to post a word of thanks again to all those that offered advice with my output problem.  Even though it was something silly that I overlooked it was great getting all those responses when I was feeling anxious about it.  A word of advice of my own from those who are still working on it.  Get mica insulators for TO-66 package transistors if you are going with the 2N4910/2N4898 to put between the PCB and the transistor.  It can save some headaches.  This compressor sounds wonderful.  Thanks to neilsk I have some Trident A-Range 10K:10K input xfmrs and I used some Carnhills I had laying around for the outputs.  The combination is incredible.  Switching between the balanced "clean" driver and the "iron" setting is not subtle at all.  It's like it sucks the room right out of the drum tracks and pulls it out front.  Pretty amazing.  I also wanted to tip my hat to Andy (sintech) and Frank (electrochronic).  Thanks for putting in the extra work to make it easier for us to get some parts.  If you haven't pulled the trigger on the meters from Andy yet, do it.  They are beautiful and they are a lot more accurate than the Retros (I tried).  Also to Fester for taking the time to churn this out.  There was a lot of controversy about this unit early on about it being a clone of another unit.  Fester's boards give the opportunity to make it in a way that is more true to an original unit, something many of us will never have the opportunity to use.  I chose to wire mine up the old school way and I'm very glad I did.  Even if this is not a "true" recreation of the old TG, it is one hell of a great sound compressor.

 

[maxwall]

Glad to see your success

First , we know that the case of the TO-66 transistor 2N4898 forms the collector connection ( collector = case) which mustbe isolated from the pads in the center where the two legs push thru( emitter and collector). there is enough tinned pad exposure on the pcb for the case to make contact and short. So its good that this is now something to be aware of.

several questions:

How are most of you finding the temperature at which the transistors are running at (2N4898 & 2N4910 specifically) ?
the orignal is mounted to the pcb with no heatsink, but improvements can be made if heat buildup is a problem. Not
sure if the original uses any type of insulator between the component and pcb.

Could the transistors be put on two short aluminum standoffs instead of a mica insulator.
reason I ask. This is how the Neve BA mic pre card does it with the 2N3055 which in effect is the same. this may or maynot
be required , just a option.

Another reason for the short standoffs is keeping heat away from the PCB  and offering a little more thermal dissipation that
may benefit the transistors service life, since these Motorola transistors are rapidly becoming expensive and hard to find.  

 

[echotodd]

The temperature buildup isn't too bad.  I've had mine on all day and I can still hold my finger to the 2N4898 which seems to get quite a bit warmer than the 2N4910.   A heatsink with aluminum standoffs would probably help the situation quit a bit and there are a few available for the TO-66 package.   At least it would eliminate the need for an insulator at the board.  To be honest, I'm surprised that more aren't running into problems by not using an insulator.  There's really not a lot of room for error when positioning the transistor so that neither of the pin 1 or 2 pads short to the shell, hence my suggestion in my previous post.

 

[bruno2000]

+1 for using 1/4" standoffs for the TO-66 transistors.
Anybody have a source for the ELMA switches?  I need 4.
Best,
Bruno2000

 

[oceres]

Hi,

Would it be possible to use Ed Anderson's EA-1166 output transformer in TG1 build?

http://hairballaudio.com/docs/1166/1166.pdf

Thanks!

 

[briomusic]

Hi,

Would it be possible to use Ed Anderson's EA-1166 output transformer in TG1 build?

http://hairballaudio.com/docs/1166/1166.pdf

Thanks!

this transformer and the VTB1148 that I am using in my TG build are based on the same vintage transformer (LO1166) and have the same impedance ratios, so I'd say - why not?