Adcom GFA-5400 bias adjustment

[vfr800fiman]

Hi all-
I just picked up a used GFA-5400, and am checking the bias adjustment. In the service manual it says:
"Prior to performing BIAS ALIGNMENT and DC OFFSET CHECK turn unit on and allow to run with rated output for approximately 5 MINUTES before attempting adjustments"
Does this mean to run the amp at full output power for 5 minutes? I'd have to get a wattage meter (or calculate the voltage across the 8 ohm power resistors on the speaker terminals).
Also, most amps I've set up I usually short the input jacks, this one says to have an input resistance of a certain value or less. Is this just putting resistors across the input jacks?
I'm a bit confused, and I want to set the bias correctly. Anyone have this amp, or similar and done this procedure?

Thanks
Glenn

 

[vfr800fiman]

Here's the full setup required for adjustments:

TEST PROCEDURES
• All tests are performed with a 115V, low-distortion (less than 2% THD), AC-power source, 8-ohm resistive load (except slew rate), and a signal source of not more than 600 ohms.
• An 80kHz low-pass filter is employed during THD distortion measurements.
• Signal-to-noise measurements are “A” weighted.
• Damping factor is measured by comparing the 1-watt output voltage with and without an 8-ohm load.
• Slew rate is measured with an inductive load, and is derived with a dual-time-based oscilloscope
reading the slope of a full power 5kHz square wave. DO NOT OPERATE THE AMPLIFIER AT FULL-POWER SINE WAVE ABOVE 22kHz OR FULL-POWER SQUARE WAVE ABOVE 5kHz.

IMPORTANT
BEFORE PROCEEDING WITH ADJUSTMENTS, MAKE SURE AMPLIFIER IS AT ROOM TEMPERATURE.

 

[so_ein_pech]

So... the test procedures are just there to tell you what they did to get the measurements in the specifications section.

You do want your amp to be warmed up before you adjust the bias. I don't know if you need to run it at the rated output for 5 minutes before you adjust the bias. I will tell you this--my GFA5300 was a bitch to bias. I would set it and then 10 min later it would be way different. Took me all day to get it to settle down. Maybe its because I just let it idle instead of running it at full power to warm it up.

BTW most people don't have watt-meters. The easiest thing to do would be to measure the RMS (AC) voltage across your load resistor in volts, square it, and divide it by the value of the load resistance in ohms.

Edit: I just realized this is a bad idea unless you have a 125+W 8Ω resistor lying around. That's a pretty effing serious amount of power for a resistor to dissipate. So, yeah, don't do that.

 

[vfr800fiman]

I suppose I should run it with some output on it to get it somewhat hot. I only have 25w power resistors on it, so as you said I'm somewhat limited to output power.

And yes, I've seen the bias float around slightly as I was waiting for it to stabilize. Time to check it again.

Thanks
Glenn

 

[mondialfan]

The only reason they tell you to run the amp at output for 5 min is to ensure the heatsinks are warmed up.

The best way to do it is to power the amp OFF and connect a set of clip on leads across the resistor you need to measure across OR even better if you have 4 clip on leads put a pair across a resistor for each channel so you can measure both channels without ever powering down. If you do not have a set of clip on test leads go and buy a set. I never recommend sticking a set of DVM probes down into the live amplifier for measuring bias. One slip of the DVM probe and that channel could be toast. I can personally attest to the fact that this can happen. I wish someone had told me this, I killed a channel in an amp about twenty years ago in just this way.

Once the clip on leads are in place sit the cover back on (don't attach it) just be careful of the leads coming out. Make sure the leads are all clear and not touching anything (shorted). Now power the amp on and let it sit for a good half hour. Be aware that if you are in your garage or workshop and the temps are way different than in the room where this amp will ultimately operate then you will see a big difference when you put it in the warmer room. If for instance it is 50 deg in your workshop where you initially adjust your bias and you bring it into a 75 deg room the bias is going to be higher once the amp is fully warmed up due to the higher temps in the house.

Ensure you have no signal coming from your pre, i.e. turn the volume to zero and select a source input with nothing connected. The best thing to do is to disconnect the amp and use RCA shorting plugs on the amps RCA inputs, but most people do not have these available so turn the volume to zero.

Now measure your bias, if either channel needs adjustment then make the adjustment, loosely sit the lid back on and let the amp stabilize again for 10 - 15 min. Each time you take the lid off the amp will cool down a bit and the bias may be slightly different than it will be once stabilized with the lid back on. If its 1 - 2mv off of what they recommend its no big deal, you'll likely never get it 100% perfect. If the temp in the room drifts a few degrees, then your bias will shift slightly as well.

 

[mondialfan]

I almost forgot. Be careful of the single turn bias pots on these amps they are very flimsy and touchy. I bought a defunct 5400 a few months back and one channel wouldn't bias up. I noticed the bias pot was pushed down and tilted back slightly. Upon closer inspection the pot had split open due to being pushed on too hard while adjusting. I put some force on the pot to force it back together and the bias would adjust properly. I ended up installing bourns sealed multi-turns in place of the cheap stock bias pots. It is now a breeze to set up the bias on it.

 

[seta1]

These amps have certain temperature sensitive components mounted to the heatsinks. You will notice that when you first power these amps up, your offset voltage may be high and as the amp warms up, it drops. This is normal. Of course, if you are also monitoring the test point for the bias voltage, it too will change as the amp warms up. The idea here is to get the amp fully warmed up before making any adjustments. 5 minutes at full rated power will heat it up pretty quick, but you can also just turn it on and play music through it at normal listening levels for around 30 minutes and it will be warmed up more then enough to make accurate measurements and adjustments for bias. This amp should be in a climate controlled room at normal room temperature of around 72 degrees when doing these adjustments. If the environment is too cold or hot, the adjustment will be off slightly. These amps do have thermal devices that somewhat compensate for environmental conditions, but only with in reason. Like plus or minus 10 degrees relative to 72 degrees. Oh, if the amp won't bias, don't be surprised of the bias pot is bad. I have run into this a lot on older amps.

 

[vfr800fiman]

Thanks mondialfan and seta1.
That's exactly what I did today, clip leads on both bias resistors with a meter on each. I loaded it with the 8 ohm wire wound resistors and played a low signal into it for about a half hour. The heatsinks were both warm to the hand, although I should have checked them with the infrared thermometer I have just to see what the temp was.

I then shorted both input jacks to make the adjustments. It was interesting to see the bias drift downward as it heated up. It finally stabilized at 35.1 mV and 34.9mV, close enough.

I then set the DC output on the speaker jacks to 0mV as specified. I could actually only get the left channel to 0.002mV, but now we're splitting hairs

I've mainly had experience building and repairing tube amps, so the SS stuff is a LOT more stable.

I've got it in my HT setup tonight, and ran it for 2 hours, still sounds great driving the Klipsch KG4's. We'll see how it does with the Magnepan MG2B's that I just rebuilt. That's the reason I bought this amp.

Thanks for the help, looks like I did it right after all. Nice to have validation though.

Glenn

 

[vfr800fiman]

I forgot to ask if anyone has a schematic for the version 2 of this amp. I got the service manual and schematic for version 1 from HiFi Engine. Apparently version 2 has some differences, mainly the speaker protection circuit.

Thanks
Glenn

 

[mondialfan]

I forgot to ask if anyone has a schematic for the version 2 of this amp. I got the service manual and schematic for version 1 from HiFi Engine. Apparently version 2 has some differences, mainly the speaker protection circuit.

Thanks
Glenn

I haven't seen one for the later 5400. I'm assuming yours has the small power supply board with the tiny transformer for the protection circuit at the front opposite side from the power switch? The best you can do is go back to hifiengine and get the 5300 V1 schematic. That schematic shows how the 555 timer protection circuit is implemented. Its an almost identical amp, only slightly lower voltage rails and 2 complementary output pairs instead of 3. The main problem I have found is that the component numbers do not match up. However at least you can trace out what is going on if you need to.

 

[vfr800fiman]

This unit does have the small power supply board in the front right corner.
That's a good idea about the 5300, I didn't know that unit had the protection circuit on version 1.

Thanks
Glenn