CA-410 Differential Pair

[Sriskie]

I have a CA-410 reading 50-60mV DC offset. I'd like it to be lower but there's no adjustment pot. I'll have to try a new matched differential pair.

Does anyone know off hand what's a good sub for 2sa872?

 

[avionic]

These will do.
http://www.digikey.com/product-detail/en/ZTX795A/ZTX795A-ND/92573

 

[Renegade]

These will do.
http://www.digikey.com/product-detail/en/ZTX795A/ZTX795A-ND/92573

Please list the originals as well as the proper replacement as there are threads here and some members I'd bet (like me) who try to keep a running list of such things for future reference.

Thanks guys, that's one of the kind of things that make AK such a great forum.

 

[petehall347]

would changing tr409 make any difference here ? just trying to understand the circuitry .

 

[Sriskie]

These will do.
http://www.digikey.com/product-detail/en/ZTX795A/ZTX795A-ND/92573

Thanks a lot.

I'll order a batch.

 

[chazix]

Just passing through, and thought it should be mentioned that the pinout of the ZTX parts is different from 2SA872. The ZTXs have the base in the middle, the 2SA872s have the collector in the middle. I have no doubt the ZTXs will work, but some extra care in lead dress might be needed. (Especially if the CA-410 has in-line through-holes for them - I don't know.)

(I'm probably behind the times, but I thought Fairchild KSA992s were the go-to choice for TO-92 PNPs with the "typical" Japanese pinout to be used as differential pairs. I'd be much interested in learning if there's something about the ZTXs that makes them a better choice. Thanks!)

Cheers,

chazix

 

[Hyperion]

I'd be much interested in learning if there's something about the ZTXs that makes them a better choice.

Higher gain probably.

 

[Sriskie]

I was actually hoping A992s would work. They're dirt cheap, I have plenty in stock, and I've used them for this purpose before. In fact, I almost threw in a pair without asking here first to see what would happen. It's a pretty basic amp anyways. At least it gave me an excuse to place an order with DigiKey and get a battery for my recently acquired Duraforce Lawn Boy.

 

[Sriskie]

Messed with it some tonight.

I matched up a pair of ZTX795A and installed them in the right channel. DC offset changed from the previous 50mV to now 68mV. Hmmm. I matched them using a Peak Atlas and they both had an initial reading of 381hfe. Subsequent readings varied depending (I guess) on how warm they were from handling and testing.

I then reinstalled the A872s and was back to 50mV. The next thing I tried was changing R444. I found this tip on another thread. R444 is a 1.8k and I replaced it with a 1.2k. That brought the DC offset down to 20mV. Better and probably good enough but I think I'm going to tweak it some more if I can find the resistors and time. I'd like to try a 1.0k next.

That seems to me like a pretty radical change. Does it make sense or am I doing something stupid? I know 50-60mV really won't make a whole lot of difference but, you know, it's part of why I have this hobby. Opinions please?

 

[petehall347]

i believe you need to match vbe not hfe . or get lucky .

 

[Sriskie]

i believe you need to match vbe not hfe . or get lucky .

Shows what I know...

Can that be done using the diode function on my DVM?

 

[avionic]

Hfe............the higher the matched Hfe the better.

 

[Sriskie]

The highest gain was able to measure out of the 10 I bought was 407. I'm don't know how the Peak Atlas measures hfe but maybe I should be looking at the individual junctions? I have a Super Cricket that can do that. Is it the B-E junction I need to match?
Also, what about changing that resistor to dial it in?

 

[petehall347]

go for matching hfe . think i was getting confused .
be sure they are all same temp. the heat from your fingers will change hfe quite a bit .

 

[avionic]

The highest gain was able to measure out of the 10 I bought was 407. I'm don't know how the Peak Atlas measures hfe but maybe I should be looking at the individual junctions? I have a Super Cricket that can do that. Is it the B-E junction I need to match?
Also, what about changing that resistor to dial it in?

Do any that you bought match ?

 

[Sriskie]

I'm not 100% sure what you're asking. Did any match with the Peak Atlas? For what it's worth, here are the readings I got the first time I measured them.

364-369-379-381-381-387-389-401-407-413

I tried the two 381s. In an effort to create a level playing field I only used the first measurement. Subsequent readings always varied. The other thing I might try is hooking the leads to each transistor and let them acclimate for 10 minutes before testing.

If you were asking if I had any that matched with the Super Cricket B-E, no, I don't know if it's a viable idea yet.

 

[avionic]

Use the 387 and 389. When I say matched. I mean matched Hfe. For example a pair that measure the same ie. 389 and 389.

 

[Sriskie]

The problem I'm having is after matching and installing the ZTX transistors it didn't get me to where I wanted to be. The offset was actually lower with the originals. I was looking for a more effective way of matching them to get the offset down but maybe the amp wasn't really expected to have 0 offset and close enough is good enough.

When messing with the resistor, I was referencing this thread.

http://www.audiokarma.org/forums/index.php?threads/dc-offset-yamaha-ca-410.242493/

Changing the R-444 resistor from a 1.8k to a 1.2k did drop the offset by about 30mV.

 

[petehall347]

what is hfe on the old transistors ?

 

[Sriskie]

That's a good question. I didn't check them or even make an effort to reinstall them the same way.