F5 DC High, Bias Low

[vdi_nenna]

What would cause the DC to go high and the bias to stay low when adjusting for the .60v across R11 and 12?

This happened after replacing the ground thermistors for diodes to help reduce a hum or changing RCAs. After a while got a sqweal in one channel and the channel never warmed up. Other side is ok.

Thanks in advance.

Vince

 

[vdi_nenna]

Think I found the problem. A bad .47 ohm resistor on R12.

I have replacements, but they are reading 1.0, 1.1 Ohms, not .47.
Even the 1% 3 watt .47 ohm reads .9 ohm.

Can I use these or will it be a problem? I'm guessing they are ok, since the extra kit I have has the exact same brand and value and they read 1.0 Ohms.

Is it best to replace them on both sides?

Thanks,

Vince

 

[anilva]

Think I found the problem. A bad .47 ohm resistor on R12.

I have replacements, but they are reading 1.0, 1.1 Ohms, not .47.
Even the 1% 3 watt .47 ohm reads .9 ohm.

Can I use these or will it be a problem? I'm guessing they are ok, since the extra kit I have has the exact same brand and value and they read 1.0 Ohms.

Is it best to replace them on both sides?

Thanks,

Vince

Generally these low value resistors measure with the error component (or lead resistance) of the dmm as well. A 0.47 ohms measuring .6 to .7 ohms is not uncommon. Best would be to have a similar resistance for R11 and R12 even if they do not measure exactly 0.47 ohms.

I don't think replacing both channels is required if the other one is good and working alright. But there are those who believe that both channels should be exactly identical.

Cheers.

 

[vdi_nenna]

Hi Anilva,

I switched both sides of the board with the 1% measuring .9 ohm. The amp is up now! Have it set to .45v across R12. Just letting it warm up. It's stable!

Thanks,

Vince

 

[vdi_nenna]

BTW- the bad ones had brittle casings after so little use. Maybe these should be 5w .47 ohm resistors. They were Mills brand. Replaced with Vishay/Dale.

 

[CanAm Man]

VDI..... what will be your final bias setting (i.e., voltage across R11 and R12)?

Recall that 0.6v across the old 0.47 ohm resistors gave you the target ~1.3 amp current flow. That won't be the case, with your 0.90 ohm resistors.

If you attempt to set 1.3 amp current flow with your 0.90 ohm resistors may also impact your limiting circuit; you may have to adjust the divider resistors (R17 and R19 for Q5, and R18 and R20 for Q6) accordingly.

You may ultimately want to go back to 0.47 ohm resistors, now that you've isolated your original problem to that resistor failure......

Just some thoughts.....

Ken

 

[vdi_nenna]

Hey Ken,

Thought you may be right, not sure anyone really has a .47 ohm resistors in their amps. These are 1% tolerance DALE/VISHAY. I have a TechDIY kit and it uses the same .47 3w 5% I used in my first F5. I didn't get a memo saying to change the bias voltage, until now. I will check the article for calculating the bias with a .9 ohm resistor, but I'm afraid that DMM leads (like stated above) add to the value and settling of said resistor may actually result in a value closer to .47 ohms. Forgive my logic, but isn't that why a pot is used to set the bias exactly?

As for the current limiter, there isn't one to worry about in this amp. That goes for the
temp compensation circuit also. Some say it sounds better without them. I think they just sound different.


In any case, the system sounds great! Crisis averted!

Thanks for your support!!!

Vince

 

[AndrewT]

I found the problem. A bad .47 ohm resistor on R12.

I have replacements, but they are reading 1.0, 1.1 Ohms, not .47.
Even the 1% 3 watt .47 ohm reads .9 ohm.

Can I use these or will it be a problem?

string together in series all your low value resistors. add a 1% low value resistor to the string.

Now pass current through the string so that no resistor is asked to dissipate more than 10% of it's maximum Pmax.

Now measure the voltage drops across every resistor and the voltage drop across the whole string. This relies on the accuracy of your voltmeter.

Remove power supply and measure the resistance of the whole string. This relies on the accuracy of your ohmmeter.

You can now work out the resistance of every resistor to ~[1% tolerance + voltmeter accuracy + ohmmeter accuracy].

If you have a 0.1%, or better, low value resistor, you can select your power resistors to a similar accuracy.

BTW,
a stock of 1% 1r0 resistors 600mW is great for building up a test resistance. Five give a 3000mW 0r2 resistance that can pass ~1.25A for ~10% Pmax and run warm, or to keep the resistors cool use a maximum of 400mA to run at ~1% of Pmax
or ~100mA for high accuracy readings with high accuracy resistors.
90mA through 0r22 gives a volts drop of 19.8mV. 900mA would take a 2000count scale of 199.9mV, to it's most accurate mVdc range. Check your multimeter tolerances. The 1.999Vdc scale is likely to be the most accurate. 10 off parallel 1r0 1% or 0.1% resistors could let you select 0r22 using the top end of the 1.999Vdc scale.

 

[jackinnj]

The 0.47R resistors are dissipating ~0.8W at idle, 1.6W peak, so 3W should be sufficiently over-rated. I used the Ohmite TWW's (5W) in the first couple of amplifiers which I built, they are great at getting the heat off the board, but folks prefer the Panasonic ERG's. In the latest incarnation I have Caddock's chassis mounted.

 

[Doug Kim]

I had same problem of high dc and low bias. My problem was k170/j74 in GR grade. I changed them to BL grade and everything was fine.