Matching Output Devices Rotel RB-991

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I bought a bunch of 2SA1492 & 2SC3856 (20 of each). My amp take 5 output pairs per channel.

My questions are :1) Is it important to match the gains to at least the Bias current on more (32mA -???)

2) The multimeter gain matching - Is it useable at all as a start for matching?

3) Is matching required? Is it because some higher gain units (out of the 5 per channel) will take on more of a load and will have a tendency to fail permaturely?

4) Is it better to match each NPN-PNP complementary pairs versus the 5 NPN's or 5 PNP's as a group?

5) Rod Elliot (ESP) had a simple circuit with Power Supply and 100 ohm POT and switch and a few power resistors (1K,100,10,1 ohm) to match the transistors at different current levels (1mA to 1 Amp). Would this be adequate for matching?

6) Is there any truth to matching the transistor to the emitter resistor? Modify the 0.22 ohm in my case to 0.15 to .33 ohm to have a very similar bias current amongst all the 5 pairs?

6) I previously matched with the multimeter (<1mA) the output devices and set the Bias at the prescribed 32 mA (7mV across 0.22 emitter resistor). The gains varied from 29 to 33 for the NPN and 39-55 for the PNP's.
But one device failed and took out its partner and took out the DC rail fuses. I had the amp idling for 8 hours to "burn it in" before I hooked up the speakers. I did have a 8 ohm 20 W resistor as the load during this 8 hour burn in.

Is this just a bad transistor or is there any effect with the possibly somewhat unmatched 5 NPN and 5 PNP's?

Any help would be appreciated!

Thanks,

Dan
 
Administrator
Joined 2004
Paid Member
Hi Dan,
Wow. So many questions.

Match all the NPN's you need, do the same for the PNP's. If you can get the NPN group's gain close to the PNP group, then great. That's what you do. This will reduce distortion. Use the same emitter resistor values for all regardless.

Match the gains around your idle current, that is where it's most important. The gains will tend to remain in the ballpark. At higher currents, the emitter resistors will enforce sharing. Your per device bias current for bipolars on a consumer amp normally runs 10~25 mA, sometimes higher but not typically over 50 mA.

One blows in an output stage, replace the lot. You can play with the "good" ones. I'd replace the driver transistors too and possibly the bias control transistor (VBE multiplier).

Make sure the heatsink surface is flat and clean, same for the transistors. Use new mica and fresh grease aplied evenly. Do not overtighten the transistor case. Work clean, clean, clean.

If you got stuck with fakes, chances are it'll blow anyway no matter how good a job you do. Buy parts from places you know sell the real thing.

I'm sorry if I ran over things you already know. I just want to cover everything.

-Chris
 
Thanks, Chris!

I believe the reason for the failure (thinking about it again) it was the fact that the bottom half of transistor was not seated right and I did not scrape the old Mica and grease off but added more grease. The legs of the transistor look like the letter Z that connects to the circuit board that is a 1/4" higher than the heatsink. So instead of scraping the old mica/grease off and adding a new mica and grease, I soldered the legs (in the Z pattern to the circuit board FIRST and then bent the legs to meet up with the heatsink. I should have secured the transistors to the heatsink (with new mica/grease) and then bent the legs to meet up with circuit board then soldered!!!

So I matched up the NPNs together and then matched the NPN's to the complementary PNP's and have probably 0.5 to 5% matching altogether.

Thanks again for the help!

Next time I will not cut corners to finish a repair.

Dan
 
Hi Dan,

The gain you meausured (on Q 6) is quite low.
According to the Datasheet, the mark on the transistor should tell you the hfe range:

*hFE Rank O(50to100), P(70to140), Y(90to180)

What hfe Rank were the stock transistors??

My experience says that you should ALWAYS prefer high hfe for the drivers and the Power trans.

The P range looks OK to me. Prefer an hfe of 100 to 110 with + - 5%

And YES you should always match ALL the transistors either NPN or PNP to the same hfe.

This will help you keep:
1. the DC offset on the output will be minimum
2. Same rise and fall times
3. better Linearity
4. Equal load on each transistor


Some people suggest that you should match the Vbe voltages between the NPN and PNP parts. Not a bad idea.
The Vbe matching will help you keep the positive and negative Clipping voltages equal.

P.S.

If i were you (as you had some power trans. failure) i would replace all the power + the drivers + the VBE multplier transistors.

John
 
Administrator
Joined 2004
Paid Member
Hi Dan,
You should be fine then. Whenever you cut corners, you may get bit. The failure to clean the old grease was a major problem and I see that more and more every day.

Hi John,
Matching all transistors simply can not be done often. Most popular numbers have the mean gain differing between NPN and PNP by 2:1. I am tickled pink when I can get numbers that will match on average between the NPN and PNP sets. Then you have junk transistors like MJ15015 and MJ15016 where NPN's are around 50~60 and the PNPs run 180~240. There is no way you will get matching there. The older 2SD424 and 2SB554 were much better, as was 2SD555 and 2SB600. You don't always want the higher hfe parts either.

-Chris
 
anatech said:


Hi John,
Matching all transistors simply can not be done often. Most popular numbers have the mean gain differing between NPN and PNP by 2:1. I am tickled pink when I can get numbers that will match on average between the NPN and PNP sets. Then you have junk transistors like MJ15015 and MJ15016 where NPN's are around 50~60 and the PNPs run 180~240. There is no way you will get matching there. The older 2SD424 and 2SB554 were much better, as was 2SD555 and 2SB600. You don't always want the higher hfe parts either.

-Chris


Well, i agree that matching is not easy, espacially with the types you mentioned!
Now, as far as the high hfe, it is convinient when the speaker is in the 3 to 4 Ohms range, or a very nasty load in general.
The driver stages will thank you for that.

This practice is good when the amp is stable and does not oscilate easily.

My favorite power devices are the MJL3281 / MJL1302.
Very easy to match them, with a good hfe curve above 100mA
 
Administrator
Joined 2004
Paid Member
Hi John,
When impedances dip like that, I normally go with a triple and use an output for a driver. You will have more outputs in parallel to handle these currents.

The newer ON Semi MJL2119X look promising. I have some to test arriving in early September. So far the MJW0281A / MJW0302A are the best for beta matches. The curves look really good on these parts.

For TO-3 parts, MJ2119X are looking good. The older MJ15022 ~ MJ15025 parts are good as well. I'm not sure about the NJL series yet. Lot's of legs in the way with those. I'm wondering if the on board diodes are worth it.

-Chris
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.