as long as it's wired correct. there will not be a problem. maka sure you massure around 0ohm across R3/R4 before you start it up🙂
i don't see any rectification ? are there some diodes hidden someplace i can't see ?
cheers Woody
cheers Woody
Don't try to recalibrate the cheap DMMs from the oscilloscope.
The oscilloscope is probably the least accurate measuring instrument you will ever use.
It's advantage is that it spreads out the inaccurate measurements over time. That advantage outweighs all the other downsides.
Compare the DMMs to each other.
Take a D cell and measure the voltage on the 1.999Vdc scale.
Reverse the leads. Does it read the same.
Repeat for the other two DMMs. What is your range of the 6 readings?
Your next step could be to buy a secondhand bench DMM. Look for >=20000count and <=0.1% on at least one voltage scale. If it happens to do both frequency and resistance then great, because these will come with increased accuracy. They will not match the DC voltage accuracy, but better than any hand held DMM. Now you can calibrate everything if you want.
The oscilloscope is probably the least accurate measuring instrument you will ever use.
It's advantage is that it spreads out the inaccurate measurements over time. That advantage outweighs all the other downsides.
Compare the DMMs to each other.
Take a D cell and measure the voltage on the 1.999Vdc scale.
Reverse the leads. Does it read the same.
Repeat for the other two DMMs. What is your range of the 6 readings?
Your next step could be to buy a secondhand bench DMM. Look for >=20000count and <=0.1% on at least one voltage scale. If it happens to do both frequency and resistance then great, because these will come with increased accuracy. They will not match the DC voltage accuracy, but better than any hand held DMM. Now you can calibrate everything if you want.
1+2,
don't set up the bias with the bulb tester in place.
Check PSU with one channel set to zero bias. Check that the bias can be increased. The bulb will start to glow. Back off the bias to zero.
Add the second channel. Check again that the bias can be adjusted until the bulb starts to glow. Back off to zero bias.
Remove the bulb tester. You have checked all the wiring. You have checked the two amps are assembled correctly. Now it's time to power up direct on line.
Set the bias of one channel, but leave it slightly low, maybe to 90%.
Set up the bias of the other channel, again to about 90%.
Leave it running for an hour or two. Check and record the output offset every ten minutes, or so.
Put on the lid. leave for another hour. Has the offset changed?
Now take off the lid and readjust the bias to final values. Put on the lid and monitor output offset for another hour.
don't set up the bias with the bulb tester in place.
Check PSU with one channel set to zero bias. Check that the bias can be increased. The bulb will start to glow. Back off the bias to zero.
Add the second channel. Check again that the bias can be adjusted until the bulb starts to glow. Back off to zero bias.
Remove the bulb tester. You have checked all the wiring. You have checked the two amps are assembled correctly. Now it's time to power up direct on line.
Set the bias of one channel, but leave it slightly low, maybe to 90%.
Set up the bias of the other channel, again to about 90%.
Leave it running for an hour or two. Check and record the output offset every ten minutes, or so.
Put on the lid. leave for another hour. Has the offset changed?
Now take off the lid and readjust the bias to final values. Put on the lid and monitor output offset for another hour.
AndrewT's above procedure is spot on.
Verify that the amp PCB has proper power voltages (+/-24v) as the first thing to check.
The important thing to remember about the F5 is that bias is a moving target until the amp is up to operating temperature. This is why you want to get it to the 90% value and wait for an hour and then remeasure. The heatsinks should be pretty darn warm by then.
If everything is looking good by that point, then set the bias to the full amount, put on the lid, and wait an hour. Check the bias again and make sure that it didn't do something weird with the top on (because it changes the overall temperature)
Verify that the amp PCB has proper power voltages (+/-24v) as the first thing to check.
The important thing to remember about the F5 is that bias is a moving target until the amp is up to operating temperature. This is why you want to get it to the 90% value and wait for an hour and then remeasure. The heatsinks should be pretty darn warm by then.
If everything is looking good by that point, then set the bias to the full amount, put on the lid, and wait an hour. Check the bias again and make sure that it didn't do something weird with the top on (because it changes the overall temperature)
The transformer, bridges and first set of capacitors are in a separate chassis.
My problem is that the bulb stays on even with bias at zero. I did verify the PSU PCB voltages were +\- 24v prior to soldering in one channel. Is it possible I had the pots turned the wrong way? Don't think so as I did verify I measured zero ohms across both r3 and r4.
I did recheck the PSU box with the amp disconnected and it still provides 24v +\- so I think I'm ok there. My next step was going to be to swap amp boards to see if the other reacts the same or not. Thoughts?
I did recheck the PSU box with the amp disconnected and it still provides 24v +\- so I think I'm ok there. My next step was going to be to swap amp boards to see if the other reacts the same or not. Thoughts?
The bulb is on bright or very dim? Does the bulb glow bright for a second after power-up and then dim? Or does it stay bright?
I'll try it tonight. Why would that work though? I think I was reading about 1000 ohms with the pots fully CCW. Then they went to zero when turned fully CW. Couldn't hurt to try I guess, just not sure what that would mean.
Measuring them in circuit doesn't mean all that much.
IF we assume that there are no shorts and the PSU is connected to the amp properly, the reason for the lightbulb tester being so bright is because you have the output Fets wide open. Run the pots to the other end of the adjustment and plug it back in. If the light still stays lit, there is something else.
IF we assume that there are no shorts and the PSU is connected to the amp properly, the reason for the lightbulb tester being so bright is because you have the output Fets wide open. Run the pots to the other end of the adjustment and plug it back in. If the light still stays lit, there is something else.
One thing to check before powering up is that the output devices are indeed electrically insulated from the heat sink. (I've accidentally cut a sil-pad and also had a flaky mica insulator....)
Where do I check this? Just from the side of the MOSFET body to the heat sink? Or from a pin?
Aha! I measured 42 ohms between the MOSFET screw and the middle pin. Same for both MOSFETs. So how is that possible? I have those super thick mica insulators separating the FET to the sink. Even if one is cracked, how would that cause a short?
So I guess I need to get a couple more of the insulators? Or is there another solution?
In the meantime, I guess I can try to setup the opposite channel. Both of those MOSFETs measured fine (open).
So I guess I need to get a couple more of the insulators? Or is there another solution?
In the meantime, I guess I can try to setup the opposite channel. Both of those MOSFETs measured fine (open).
Hmm, I removed the screw entirely and still get 44 ohms from the center pin to the heat sink. So not sure a shouldered washer would help here. Is that logic sound?
If you follow the schematic, you will see the drains (middle pin) going to ground through R1/R2 and R5/R6 (at least), so it should read something (when the hs is grounded). But that >something< should be the same on all output devices.
- Status
- Not open for further replies.