gar2520 build thread

[3nity]

CM75101APC only has 2 ratios 1:5 or 1:10.
I've always used a 150K on the sec. load.
Thats the value everyone uses....

 

[mikefatom]

I'm using the 600 ohm input.

 

[mikefatom]

I would be pretty concerned if changing the load to 15k solved my problem.
Everyone uses that load value as 3nity mentioned without problems. 15K is far from 150K...

I've checked and rechecked everything! I'm stumped!  ???

 

[mitsos]

Mike transformer Ratios are 1:5 or 1:10 nothing else!!

the 1:8 I was talking about is not a Cinemag 75101.  But, yeah I should have maybe stuck with the OPs components in the example.

With a 1:5 input, 1:2 output, and standar 40dB opamp gain, you have 60dB exactly, so maybe that's the way to go.

Re: load Resistor, you want it to give you 1K5 on the input side.  
For a 1:10 you use 150K because  150K / 100 (impedance ratio = turns ratio squared) = 1K5 input impedance.

For 1:5, (ad 600 Ohm impedance) if you work backwards off the 600 Ohms input, then multiply by 25 (impedance ratio) , you get 15K as your load R. Maybe the others can chime in?   So sounds like you found your problem?? let us know..

Anyway, with the Cinemag, connect pins 2 and 3 to put the windings in series.  

I went through these same steps 2 weeks ago and it ended up being the diode, that's why I mentioned it. good luck.

 

[jsteiger]

I'm using the 600 ohm input.

Mike,

The 600 ohm input will be 1:5. Try rewiring for the 150 ohm and step up of 1:10. Ed's EA2622 are about 1:7 to one. He says that is pushing it with a 2520 style DOA. The 600 ohm 1:3.5 is basically unusable. I believe that is why API changed to the 1:8 and then the 1:10 later on. Some of the common input trafo's are internally wired for the higher step up with no other option.

Someone correct me if I'm wrong.

150K is my load R as well.

Jeff

 

[mikefatom]

Well, here is what solved my problem.

I wired for 150 ohms (1:10 ratio) and it made the problem even worse!
Then, I strapped a 20K resistor in parallel with the 150K RL (approximating to 17K)
and all high gain craziness went away!

can someone explain. I suck at transformers...

Thanks to silentarts, 3nity, mitsos, jeff and all!!! you guys rock!

cheers,

mike

 

[clintrubber]

can someone explain. I suck at transformers...

We've seen many strange things here, but this is the limit  8)

Maybe go see a doctor ?




Sorry 

 

[mikefatom]

lol... 5 years of engineering school and I already forgot 75% of it!

can anyone confirm that they have used the 150 ohm wiring with RL=150K successfully at all gains?


thanks

mike

 

[[silent:arts]]

... can anyone confirm that they have used the 150 ohm wiring with RL=150K successfully at all gains?

yes. but with another input transformer (EA2622) in slightly different schematic / PCB (Classic VP26)

 

[mitsos]

Mike,

I've used both, though I don't remember what RL I used with the trafo 1:5, but I am pretty sure it was a lot lower than 150K.  For 1:10 I always used 150K.

But it's not really clear: you mean you used the RL=17K with the trafo as 1:5 or 1:10?

good to hear it's cleared up!

 

[mikefatom]

hey mitsos,

I actually mean, that the only configuration that worked perfectly for me is,

Input TX : CM75101APC wired for 150 ohm input.
RL = 17K.

does that all sound normal?? It sounds great here, but just wondering...

cheers

mike

 

[jsteiger]

hey mitsos,

I actually mean, that the only configuration that worked perfectly for me is,

Input TX : CM75101APC wired for 150 ohm input.
RL = 17K.

does that all sound normal?? It sounds great here, but just wondering...

cheers

mike

Mike,

My pre circuit has a 220pf C and a 10K R in the Zobel network, a 150K LR and a 1K series R. So, a little over 10K I gather. That is for a vintage AP2622 or one of Ed's 2622's. Looks like you are right in the ball park.

Like Volker said a little while back...

...try 15K...

Look where you ended up! Pretty close. 

Good to hear you got it all worked out.

Jeff

 

[mikefatom]

Thanks Jeff, Volker was indeed right! thank you Volker!!!

I was just making sure I wasn't misusing the TX.

cheers

mike

 

[jsteiger]

Hey all,

Here is a link to a post in my VP2x support thread. Volker has ran about a dozen Audio Precision tests of my pre with many different DOA's in it. He also has a reference of a vintage BAE racked 312 card.

Very interesting. I have to say the gar2520 fares very respectably in these tests.

http://www.groupdiy.com/index.php?topic=33501.msg418425#msg418425

Cheers, Jeff

 

[diko2]

Hi all,

I've just finish 2 gar2520 today. (the assembly manual is really clear by the way...)
I use them with reliquia 312 V3
They seems to work fine but, in both of them, the BD140 and the BD139 are really hot ( can't keep a finger on it...)
Is this normal?
They run on +18v/-18v.

Cheers

Diko

 

[mikefatom]

Hey diko2,

Just to give you a bit of perspective of the BD transistor's temperature. Running them at 18V rails do get them hot but I've measured a little less than 20mA through them at that voltage! Since they are rated for much more than that (1.5A I believe), I don't think it's anything to get worried about.
Try running them at 16V rails. They get cooler and still provide great headroom.

My BD139/140 based 2520s also run toasty, although not as much. At least it's my experience.


btw, could you tell us what values you're using for your RL, Rf, Cf and which TX input ratio?

cheers,

mike

 

[diko2]

Hi Mike,

My 312 use

-Cinemag CMMI-8PCA for input
-RL=100k
-RC=9,76k
-CC=22p

Thanks

Diko

 

[gar381]

Hi all,

I've just finish 2 gar2520 today. (the assembly manual is really clear by the way...)
I use them with reliquia 312 V3
They seems to work fine but, in both of them, the BD140 and the BD139 are really hot ( can't keep a finger on it...)
Is this normal?
They run on +18v/-18v.

Cheers

Diko

Diko +/- 18 volts is fine.  I use a pair of 553 EQ as a test bed for my gar2520 DOAs.
I burn in my assembled ones at +/-18v in my 553s. Yes they do run a bit hotter than +/-15v
I can only touch the +/-18v BD139/140s for a few seconds without discomfort. 
At +/-15v they are only a little warm. If you have concerns about heat back down the PSU voltage.
The headroom difference between +/-18v and +/-15v is less than a 1 db.  Lots and lots of Headroom
either way.  Hopes this helps..

GARY

 

[jsteiger]

I broke in a few for myself, testing my DOA assemblies this past weekend. I had them running in my desk which is on the API standard, +16V/-16V's. They got a little warm but nothing to melt the skin.

Can you guys imagine the stress put on an epoxy potted amp running at 18V or even 20V? Little to no heat dissipation. ??? I really feel good about the open, unpotted design.

As far as Huntington's, if my board is under power for a few hours, some of the vintage ones are barely even warm to the touch. The, some are very warm, even too hot to handle. Great song

BTW Gary, I love this build. It's a great satisfying feeling to ease thru the assembly, doubling checking my dumb-ass self for errors, then fire one up and have it work straight off the bat...no issues!

Cheers, Jeff

 

[gemini86]

Can you guys imagine the stress put on an epoxy potted amp running at 18V or even 20V? Little to no heat dissipation. ??? I really feel good about the open, unpotted design.

Isn't the potting supposed to act like a radiator, dispersing the heat across a larger surface, allowing it to dissipate faster? Shout me down if I'm wrong... I may be miss-informed.