Pioneer SX-450 DC Offset

Bubba57

Active Member
I have a little SX-450 here that had 2 blown outputs. I replaced all 4 with MJE15032G,s and I also replaced Q201-Q204 with KSA992,s. No caps were replaced. I matched up Q201 and Q203 to within a few points of each other. There was a little larger spread between Q202 and Q204, but still less than 10 points.

I now have a DC offset of ranging from -18mV to -23mV on the Left Channel. The offset on the Right channel is ranging from -16mV to -20mV.

The idle current for the Left channel is running 52.5mV. the Right side is at 61.6mV.

What would cause the DC offset to be so far negative, and the idle current to be so high?

Thanks for your help
Bubba57
 
Is the jumper in the instructions cut or intact? All 4 transistors should be matched up to each other (Q1 thru Q4) not just in pairs(Q1-Q2 and Q3-Q4).

You did attach the probes as outlined in the manual, right? Offset looks ok. Don't think the + or - makes a difference. (see this thread from Echowars
" Amplifier Distortion, DC OFFSET, and YOU " )

With no pot to adjust the idle current it's set by resistor(s) R123 and R124, with the shunt (jumper) across it. It's a real basic set-up but PIONEER had been using it from at least the SX-434 maybe earlier.

Also the MJ's might have something to do with the idle current and or offset. OnSemi's tend to have a lower turn-on voltage than most other brands. Mark the Fixer or Echowars can explain how it affects operation.
 
Thanks Larry for your response. The jumper is still intact and I measured the idle current at the pins per the manual. It (the manual) is rather vague as it says if the idle current is less than 15mV, cut the jumper. It doesn't say what to do with 50 to 60mV.

I did read that link you attached, a couple weeks ago I believe it was. Hopefully Echowars will correct me if i'm wrong, but I took it mean maybe a few negative volts was Ok. I wasn't sure about -20mV. Both channels are about the same, so my guess is that the R123 and R124 are probably OK. But then again, I also guess that with the jumpers intact, it bypasses these resistors??? So, zero resistance (jumpers intact) would yield a lower idle current.

Does the idle current value, actually have any affect on the DC voltage at the speaker terminals (being it pos or neg), or is it just another health check?

These numbers may be normal for this 450, but I didn't want to hook it to any speakers without checking from the AK group first.

Thanks again Larry
Bubba57
 
Idle current and DC offset are two separate issues.

The best way to match the diff pairs is to leave one a992 in place and swap other 992's in and out of the other position of the pair - looking for the lowest DC offset.

BE SURE the power supply voltages have decayed before pulling out any a992.

By using the mje's, you disturbed some idle current assumptions. ON-semi's slightly lower Vbe turnon results in higher idle currents for a given stv diode voltage.

With how small the trim pots are, you COULD shunt a 1000 ohm single turn trim pot (652-3386H-1-102LF) across for R221 (r222) (1000 || 82 = 76 ohms)
or substitute R221 (r222) with a 100 ohm trim pot (652-3386H-1-101LF)

THEN idle current is adjustable. Remember to tie the center tap and high end together as this is a variable resistance and not a potentiometer tap.

The idea of shunting provides an "emergency fallback" if the trim pot goes bad or is mis-wired.


It's just that the price point of this BOTL model didn't include the luxury and expense of a trim pot.
 
Last edited:
I appreciate it Mark,
Some of this echoes what Larry was saying. I think I will, add a couple of VR,s just to see if I can bring the idle down. It's worth a shot. And thanks for the education.

Bubba57
 
So I didn't understand, is -20mV DC offset anything to be concerned with? I know, at least I think I do, that a positive DC will pulse the woofer forward. Wouldn't a negative DC tend to pulse the woofer backwards or even cause a enough resistance to hold it back from its intended/commanded full forward travel?

I'm thinking along the lines of a subwoofer, where you flip the switch from normal to reverse.

Bubba57
 
Mark's procedure matches the KSA992's for Vbe. I bet you were matching them for hFE, which is less critical here. In the absence of a trimpot, the Vbe matching between Q201 and Q203 [Q202 and Q204] controls left
channel DC offset. Leaky C207 or C219 [C208 or C220] will also cause offset.

With perfectly matched parts, this power amp design will still have a small offset, due to Q201's [Q202's] base bias current. The coupling cap can make it more positive, the bypass cap can make it more of whatever polarity it already is. Since your offset is negative, Vbe mismatch is the dominant player.

If you install idle-current pots, use 100 ohms in place of the resistor. 1000 shunted across will compress the meaningful adjustment range into the last few percent of rotation.​
 
Last edited:
You boys are GOOD !!! Dave, I was in fact matching the hFE values on the A992's. If you recall my post (and you answered ) about using the MJE outputs, I did use them in all 4 positions. Probably for the reasons both you and Mark talked about,that resulted in a high (50 to 60mV) idle current in both channels. Per Marks suggestion, I installed two pots to bring the idle current down. I did deviate slightly though. I determined that in order to bring the idle current down to 20mV (the spec for my SX-950) instead of the 82 ohm in R221 and R222, I would need something like 56 ohms on one channel and 67 pn the other. I decided to just pull the resistors and put 2 pots in their place. I only had 1k pots. Again you are right, the final adjustment is closer to the extreme of these pots, but I was able to dial them in at exactly 20 mV.. I don't even know if that is a good value for this little thing.. So the MJE,s did create high idle, the pots compensated for that, the 100 ohm would have been better suited. I still have -20mV offset on one channel a -17mV on the other. The numbers are fairly close side to side. Do you still think matching the vibe of Q201-204 will make much of a difference? I'm not concerned about a few mV variance, side to side. I just want to the DC negative voltage for both channels to be closer to zero. I don't know what would drive the offset so far negative but I'm not comfortable with it.. Again it is both channels so even if it was the caps you mentioned, it would have to be all 4 wouldn't it? I will definitely check these caps and play around with these trans.. Could the characteristics of the MJEs be causing the -20mV dc offset? If so, I think I have 4 of the 2073's I could throw in..

Like I said y'all are good..
Thanks for the help and I would just like to get yours an Marks thoughts on this. Maybe I could even add trim pots to the dc offset

Bubba57
 
With 19,000 plus posts over 10 years here, absolutely.

I have in the past named exact parts to replace to fix a specific problem, after tests done without even opening the bonnet. Correctly.
 
Well friend, you did it one more time, absolutely and correctly. Not only did you not open the bonnet, in this case you're a couple thousand miles away from the victim, I mean patient.

Thank you Mark
Bubba57
 
With how small the trim pots are, you COULD shunt a 1000 ohm single turn trim pot (652-3386H-1-102LF) across for R221 (r222) (1000 || 82 = 76 ohms)
or substitute R221 (r222) with a 100 ohm trim pot (652-3386H-1-101LF)

THEN idle current is adjustable. Remember to tie the center tap and high end together as this is a variable resistance and not a potentiometer tap.

Putting in the 100 ohm trim pot worked perfectly! I can now trim the voltage between posts 19 & 20 and posts 17 & 18. Excellent!

Now I've got two questions:

1. The Pioneer SX-450 service manual (page 28, titled 9.4 POWER AMPLIFIER SECTION) refers to this as IDLE CURRENT, but this thread and the comments around a trim pot all sound like DC OFFSET. Which am I adjusting when I adjust the trim pots I've put in place of the old resistors at R221 and R222?

2. Using the 100 ohm trim pots I can crank the measurable voltage between posts 19 & 20 and posts 17 & 18 all the way down to 0.01mV. What is a good target for this? 10mV? 5mV? What is a great target? 1mV? 0mV?
 
if you set idles at 15mv you are in the safe zone .. it may however produce better sound at 20mv . and still be in safe zone .
 
This thread is gold, I just did the same trick with the 100ohm trimpot.
I was able to bring down the idle current from 120Mv to 20Mv.
Great!
 
Same here, thank you for all information! With trimpots at R221 and R222 i set the idle current back to 25Mv instead of +60Mv.
 
Back
Top Bottom