Calculating Current Through a Transistor

[LBPete]

I need to replace the 2SC1583 NPN dual junction transistors used on the driver boards of a Sherwood S9910 receiver. I want to use a pair of KSC1845 transistors. These transistors are rated at half the collector current of the 2SC1583 so I want to verify they are suitable replacements.

Looking at the schematic the current flow appears to go from a 58V source through a pair of 3k resistors (R603/604) to the collectors of the dual junction transistor, On the emitter side it passes through a 10k resistor (R605) then to the collector of TR602 and through the emitter to a 1.2k resistor (R606) and to ground.

Looking at a single junction of the 2SC1583, the total circuit resistance is 14.2k which calculates to a little over 4mA. The driver board schematic is attached. Am I looking at this correctly?

 

[rcs16]

You can figure out the current, they post voltage readings
the current thru each device is 0.5mA 1/2 of (-10.6- (-0.6)) / 10K, R605
ksc1845 are fine, match them by measuring Vbe ( using ohmmeter in diode mode) and match a pair that measure within 1mV

 

[elnaldo]

If you don't have a easy way to match the transistors, a very easy method is to solder them directly on the board and measure DC offset. You can solder on the traces side of the PCB to do it easier, less PCB damage. If too high, replace just one of them until you find a pair that gives you a low DC offset (I'd say less than 20mV). Once found a pair, install properly but don't swap them, leave each one at the same position you already tested.

 

[LBPete]

The reason I need to replace the C1583 is the DC offset is extremely unstable and heat sensitive in one channel. This receiver doesn't have a DC offset trimmer so it's very dependent on the differential pair.
I just picked up a Peak DCA55 for matching. If anyone is interested Newark has by far the cheapest price for them.

You can figure out the current, they post voltage readings
the current thru each device is 0.5mA 1/2 of (-10.6- (-0.6)) / 10K, R605
ksc1845 are fine, match them by measuring Vbe ( using ohmmeter in diode mode) and match a pair that measure within 1mV

I'm not following your calculation. I can do the math but where does the (-10.6- (-0.6)) come from?

- Pete

 

[gadget73]

0.6v is typical voltage drop across a silicon transistor, so you take your measured 10.6, subtract the standard 0.6v, and its actually dropping 10v from current.

 

[LBPete]

OK, standard voltage drop is -0.6 but where does the -10.6 come from? The schematic shows the applied voltage as 57v.

- Pete

 

[gadget73]

That I don't know. Normally I'd just measure the voltage drop across the collector resistor and use that and the resistor value to work out the current flow. Or you do the same across the emitter resistor and divide by 2 since its feeding 2 devices. Thats what I'd do on a tube anyway, I presume transistors behave close enough for the theory to hold.



**edit**

I see. Its 10.6v on the bottom side of the emitter resistor and 0.6v at the top of the emitter resistor, so 10v drop across the 10k emitter resistor. Thats 1ma for both devices, or 0.5ma per-device.

 

[LBPete]

Oooh, I see it now. Those values are on the schematic.

- Pete

 

[LBPete]

I was able to exactly match 4 C1845s and replaced the C1583s in both driver boards. The left channel had zero to 1mV DC offset before and maintained that after the swap. The right channel had high and unstable DC offset before the swap and a very stable 20mV after the swap. It corrected the problem but I'd like to see lower DC offset for that channel. There is no trimmer adjustment for DC offset on these boards so I'm guessing there is some variation in fixed resistors between the two boards in that circuit.

As a side note, this receiver is a joy to work on. Everything is accessible. One thing that is odd though is the Bias Adjustment procedure. It requires a distortion analyzer, audio signal generator, watt meter and an oscilloscope. I've measured the voltage drop across the emitter resistors and set it to 25mV. That should be pretty close.

- Pete

 

[rothwellaudio]

The right channel had high and unstable DC offset before the swap and a very stable 20mV after the swap. It corrected the problem but I'd like to see lower DC offset for that channel.

There's been some misleading stuff written about DC offset with the implication that perfectly matched transistors in the differential amp will give you zero DC offset. That's not quite true.
The two transistors of the long tail pair each pass base current through a resistor, and that current generates a voltage across the resistor. If the resistors aren't matched, or if there's leakage across the large DC blocking capacitor in the feedback network, there will be a DC offset even with perfectly matched transistors due to the voltage across the two resistors not being equal. At the design stage the offset can be minimised by selecting low values for both resistors (to minimise the voltage across them), but the price to pay is a lower input impedance and a higher value of DC blocking cap.
Without going to the lengths of a redesign you could try using low leakage caps in the feedback network and matching the relevant resistors.

EDIT:
Having looked again at the circuit diagram, I'm a bit confused. On one side of the LPT there's 103.3k to ground (R601 + R602) and on the other side there's 47k to the output (R612). I would expect these two resistances to be the same to maintain zero offset. There must be something else in the circuit I've overlooked.

 

[Binkman]

There's been some misleading stuff written about DC offset with the implication that perfectly matched transistors in the differential amp will give you zero DC offset. That's not quite true.

Quite true..
((Aside from potential schema errors, if those resistors are as noted above check out. Just swap that pair on the low dcmv side to the other and see. But getting a perfect pair is like a needle in hay stack over lot numbers.. some buy 100's to test/match.

 

[elnaldo]

I had good results swapping the 2 transistors in the pair, sometimes they work better in one position . Also, replacing just one of them, and swapping positions if the offset goes higher.

Anyway 20mV is very low, don't worry too much. I wouldn't risk the unit or the PCB to correct this small difference.

In cases of larger offset (100mV or more) I had success modifying some resistors in the network (usually soldering a large resistor in parallel to reduce the resistance just a bit, search a thread called "DC offset puzzle"). For example, you could try a 100K resistor in parallel to R603 or 604 and check if the offset goes up or down. Also, replacing some resistor with a trimpot could adjust this small difference.

 

[rothwellaudio]

Anyway 20mV is very low, don't worry too much. I wouldn't risk the unit or the PCB to correct this small difference.

I agree. There's always a chance of lifting a pad or track when removing components and I advise against doing stuff that doesn't need to be done.
BTW, I knew a guy that was obsessed with the idea of improving his amp and was always wanting to modify and change things. He'd replaced parts so many times that there were hardly any pads left on the board and he had to try to attach bits of wire here and there to make connections and it ended up being a total mess. If it ain't broke, don't fix it

 

[gadget73]

Check the 3K emitter resistors to verify they are on value. Drift there will skew things a bit.

but I tend to agree about 20mv being not worth fussing over.

 

[LBPete]

This thing is driving me crazy. I replaced R603 and 604 with matched pairs of 3k resistors and it did seem to drop the DC offset but before long it was up to it's original tricks. Very high DC offset and crackling form that channel on start up. After playing for awhile it drops below 20mV but occasionally spikes back up. I replaced the differential pair a second time and for good measure replace all the other transistors on both driver boards. One board works perfectly the other board continues with the high DC offset and crackling on start up then dropping down after playing for 5 minutes or so.

All the electrolytic capacitors were replaced and I used film caps for the small values. I've checked the ceramic caps for shorts using an ohmmeter and compared the results left and right. I've compared the resistance values for all the resistors and compared the values between the good and bad boards. I did the same for the diodes and there are no discrepancies between the two boards. The problem follows the board when switching boards between channels. The good board is always good and the bad board is always bad.

What am I missing?


- Pete

 

[gadget73]

You may need to monitor voltages between the diff pair and the output to see if you can find the spot where things start acting up. I'd expect to see spikes on the meter or scope that go along with the offset shift at the speaker terminals. Its possible that a failing part somewhere later in the chain is causing the voltage drift and the noise. I've had transistors get leaky and make those sounds. Possible freeze spray and/or heat will help chase it out.

 

[petehall347]

have you narrowed it down to driver board ?

 

[sregor]

Try changing d601 and d602. D602 should be equivalent to 1n4148. D601 is a 6.2 volt Zener. They control the voltages for the current sources, and could be problem. Magnifying glass and check for cracks and bad connections, but I assume you've already done that. my 2 cents.

 

[LBPete]

You may need to monitor voltages between the diff pair and the output to see if you can find the spot where things start acting up. I'd expect to see spikes on the meter or scope that go along with the offset shift at the speaker terminals. Its possible that a failing part somewhere later in the chain is causing the voltage drift and the noise. I've had transistors get leaky and make those sounds. Possible freeze spray and/or heat will help chase it out.

The issue clears up withing a few minutes and the voltage spike induced pops only occur for seconds after the speaker relay turns on. I'll monitor both channels to see if voltage fluctuation occur at the same time on both boards. I already tried freeze spray on the transistors. I'll try it on other components today.

have you narrowed it down to driver board ?

Yes. This receiver has separate driver boards for each channel and the problem follows the board when the two boards are swapped left to right.

Try changing d601 and d602. D602 should be equivalent to 1n4148. D601 is a 6.2 volt Zener. They control the voltages for the current sources, and could be problem. Magnifying glass and check for cracks and bad connections, but I assume you've already done that. my 2 cents.

I've checked those diodes on the board but not the zener. I've also looked at the foil but at this point it couldn't hurt to look again.

Thanks for the suggestions.

- Pete

 

[LBPete]

Still no resolution. I monitored the emitters of the differential pair and can see the spikes there. Then I monitored both transistor bases and can see the spikes there as well. It's kind of difficult to see because the spikes only occur on initial start up after sitting unused for hours. When they occur both o'scope traces jump up but the spike appears to occur more on the base of the input side of the diff pair. That side is just coupling caps, a resistor and a little 100p ceramic cap. The coupling cap was replaced with a Wima film cap and I swapped out the resistor and the ceramic cap. I also replaced the three diodes on the board for good measure and sprayed all the components with freeze spray. But it still occurs.

Part of the problem chasing this issue is it goes away quickly and requires a prolonged cold soak to get it to occur again. When switched on, the DC offset is high, over 1.5 volts but it comes out of protection after a normal delay. If popping occurs at all, its only two or three pops and only in the first few seconds after power on. When those initial pops occur the DBT surges. The DC offset drops quickly and steadily and within a few minutes it drops to less than 10mV. At that point you can play it all day without incident and it sounds great.

I'm starting to get a little concerned about the protection function. I've never seen an amp that will come out of protection with over 1.5 volts on the output.

- Pete