So I traded one my amps for a Dynalo that was built by another person. The previous owner did not even build it, but that is another story.
I was trying to verify some voltages and must have touched or shorted some leads, and a small poof of smoke arose. It remained on, but I now had 14mV DC offset on the output instead of the normal about 1mV. In taking the head sink off that was glued on, I broke a couple of the transistors (heads popped right off those TO-92's).
Long story short is that I have a Dyanlo with broken transistors and the Toshiba 2SA1015 and 2SC1815. I could get GR replacements from AMB, but those are $0.60 a piece. Mouser sells the Y gain ON Semiconductor replacements that are $0.14 a piece.
I already bought 20 of each PNP and NPN (40 total) of the ON Semi Y ones from Mouser. My plan is to match them as close as I can.
My question is...as I know the there is an HFE gain difference between the Y (120-240) and the GR series (200-400), what is the AUDIO difference between these two?
Installed were the Toshiba GR's, but will be replaced by ON Semi Y's.
I was trying to verify some voltages and must have touched or shorted some leads, and a small poof of smoke arose. It remained on, but I now had 14mV DC offset on the output instead of the normal about 1mV. In taking the head sink off that was glued on, I broke a couple of the transistors (heads popped right off those TO-92's).
Long story short is that I have a Dyanlo with broken transistors and the Toshiba 2SA1015 and 2SC1815. I could get GR replacements from AMB, but those are $0.60 a piece. Mouser sells the Y gain ON Semiconductor replacements that are $0.14 a piece.
I already bought 20 of each PNP and NPN (40 total) of the ON Semi Y ones from Mouser. My plan is to match them as close as I can.
My question is...as I know the there is an HFE gain difference between the Y (120-240) and the GR series (200-400), what is the AUDIO difference between these two?
Installed were the Toshiba GR's, but will be replaced by ON Semi Y's.
So even if the originals have an hfe of say around 280 (for example), and the newer ones I match to around 200 or so, or even less, say 145, how much will actually change, if any?
Meaning, will there be any change in actual gain? More noise? etc.
There will be no apparent change in final gain.
The higher gain devices would produce an amplifier of higher Open Loop Gain, that is to say the gain without any feedback wrapped around it. Once the loop is closed the gain as far as you are concerned will be the same.
Differences in the final noise figure will similarly be microscopic. Maybe better, maybe worse, but its a quarter of a half of nothing difference.
The higher gain devices would produce an amplifier of higher Open Loop Gain, that is to say the gain without any feedback wrapped around it. Once the loop is closed the gain as far as you are concerned will be the same.
Differences in the final noise figure will similarly be microscopic. Maybe better, maybe worse, but its a quarter of a half of nothing difference.
Thanks. Certainly there is closed loop gain in the Dynalo. So thank you.
Thanks. It might actually be quite welcome then, I found the GR's kind of warm. The amp had low end slam, but not a lot of bass texture and definition.
I guess I will find out soon enough. Looks like I don't have much to worry about then.
I measured few of the Toshiba's that made it out of my board safely, it seems they are around 200 hfe. The On Semi's are on the way.
You have to be careful with those. The pin out is different than the 2SC1815. Most other specs the same, although hfe gain is about double and also much higher bandwidth.
Thanks for your reminder. Actually I have modified the schematic and pcb, shouldn't have any problem on pinout. But I have a question regarding to matching. Supposed I have limited number of BJT to match, what's the priority to match the devices? Should I match (Q1, Q2) then (Q3, Q4)? Or (Q3, Q4) first then (Q1, Q2)?
From what I read in other forums, Q1/Q2 matching and Q3/Q4 matching helps keep the DC offset low. There is only so much adjustment on the pot, and then the active servo does the rest (up to 5mV). The LCD forward voltage matching also helps. But not too much worry on performance.
Matching the parallel buffer transistors helps keep even load across all of them, but the emitter resistor is pretty high, so there is not much worry here, especially if you have a shared heat sink.
My KSC1815's I got from Mouser were pretty well matched and close. Only a few out of the main grouping. Many in the 150 hfe range. However the KSA1015Y's were low and high. 135 on the low grouping, and a few in the 170, 180 range. And I bought 20 each, only needing 12 each. I'm just going to get as close as I can. I am also using a cheap hfe meter on my DMM.
Oh, and my original Toshiba 2SC1815GR's were around 200. So not bad the On Semi Y's being in the 150 range.
Well, when I check the hFE on my DMM, the BC550C is always higher that BC560C. I can only found one or two out of my 15 pieces closely matched with BC550C. It is hard to match when the quantity is not large enough.
BTW, when I check with the datasheet, the BC pairs have two parameters, hFE and hfe(small signal). I find that my measurement using DMM(4XX) is very different from using the method provided by AMB(7XX) (The β22 stereo amplifier), anyone knows what's going wrong? Should I match hFE or hfe??
BTW, when I check with the datasheet, the BC pairs have two parameters, hFE and hfe(small signal). I find that my measurement using DMM(4XX) is very different from using the method provided by AMB(7XX) (The β22 stereo amplifier), anyone knows what's going wrong? Should I match hFE or hfe??
I don't think there is a difference in hfe and hFE. I do know DMM only checks small small mV insteat of 12V in the AMD description. The AMD matching is fine, actually measuring the amps directly, but you can also put a resistor on the collector and an emitter side and measure the volts across the resistor across the resistors and calculating hfe.
Transistor matching
Values of hfe will be slightly different using real world voltages.
And I have no idea why NPN's measure higher.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.