afternoon everyone
i need a bit of help setting up the idle current on my old teac dc servo A9 amp, it has how to do it in the schematic, but it involves some serious test equipment! a AF ocsillator, an attenutaor, distortion meter and.. an oscilloscope!
now just being an odd diy'er i dont have this gear, ive never tried to do this before and am no expert, all ive built is a few cmoys, a phono stage and a power supply, as well as recapped a few old amps, my electronics experience is very minimal.
i came across this post on TNT audio:
Adjusting the Bias Of Your Amp [English]
but i dont think im understanding the full jist of this, when its stated:
"Then, put one wire on one side and the other on the other side of an emitter resistor, noting its value. Say it is 0.22 Ohms and say you have a voltmeter reading of 10 mV. The quiescent current is obtained by dividing the voltage with the resistance, i.e. 0.01:0.22=45.45 mA. "
so would that mean shove a high wattage reistor (8 ohm?) at the speaker output terminal then measure the voltage from one side of the resistor to another? its just in the teac schematic it shows the resistors as partially grounding the speaker signal, the bulk of it going to the test equipment.
and would this method even work for a dc servo amplifier? and if so what ideal current value would i be looking for?
many thanks
i need a bit of help setting up the idle current on my old teac dc servo A9 amp, it has how to do it in the schematic, but it involves some serious test equipment! a AF ocsillator, an attenutaor, distortion meter and.. an oscilloscope!
now just being an odd diy'er i dont have this gear, ive never tried to do this before and am no expert, all ive built is a few cmoys, a phono stage and a power supply, as well as recapped a few old amps, my electronics experience is very minimal.
i came across this post on TNT audio:
Adjusting the Bias Of Your Amp [English]
but i dont think im understanding the full jist of this, when its stated:
"Then, put one wire on one side and the other on the other side of an emitter resistor, noting its value. Say it is 0.22 Ohms and say you have a voltmeter reading of 10 mV. The quiescent current is obtained by dividing the voltage with the resistance, i.e. 0.01:0.22=45.45 mA. "
so would that mean shove a high wattage reistor (8 ohm?) at the speaker output terminal then measure the voltage from one side of the resistor to another? its just in the teac schematic it shows the resistors as partially grounding the speaker signal, the bulk of it going to the test equipment.
and would this method even work for a dc servo amplifier? and if so what ideal current value would i be looking for?
many thanks
Dont connect a speaker in case there is a DC offset to confuse matters.
Just put your meter across the 0r22 set on mV range.
Adjust the bais to the meter reads 10mV.
No it doesn't mean that at all. I would leave well alone as you can do more damage than good messing about with the bias. Unless you have replaced output transistors etc there is no need to reset the bias.
The Teac manual is on about setting it the correct way by monitoring distortion with test equipment. The TNT site is bad information.
thats the problem richie i have replaced 4 transistors in the power amp section that were fried, the amp works, but i would like to get it set up to spec as its not sounding as good as it did before.
i tried taking it to a few audio repair shops, none of them will touch it , one guy even said "look for a grey haired old guy, cos theyre the only ones who will touch it..." really helpfull..
What voltage do you get across the 0r22 ?
If its much above 10mV then you might end up damaging your amp again.
If its a lot less then you will be getting crossover distortion.
havent even got that far yet nigel as i was unsure how to set it up,but i'll have a trip to maplin and get a high wattage 0.22ohm resistor
i must admit i didnt really understand the TNT post saying to ramp it upto 100ma no matter what the amp, wouldnt that be biasing it towards class A and screw up the components?
i must admit i didnt really understand the TNT post saying to ramp it upto 100ma no matter what the amp, wouldnt that be biasing it towards class A and screw up the components?
No, wait. You don't need extra resistors.
The idea is to measure the voltage across resistors that are already inside the amp.
here we go:
Download Teac_A7_A9.pdf from Sendspace.com - send big files the easy way
thanks for the help so far everyone
Download Teac_A7_A9.pdf from Sendspace.com - send big files the easy way
thanks for the help so far everyone
Right on, that TNT site is a bit misleading. The whole idea is to maintain a constant Gm vs current across the zero current crossing. The amount of current for optimal bias depends on the type of output devices, Zout, and the transconductance curve. Too little or too much bias can increase distortion in a class AB follower circuit, MOS or BJT.
Hi,
the part you need is on the last page.
It shows a two stage EF output.
The emitter resistors are 0r22.
Optimal output bias will be ~20 to 25mV across each emitter resistor.
Short the input of the amp (this should have the same effect as turning volume control to minimum).
Remove the speaker cables.
Measure the DC voltage across Re. It should be mV not volts and not zero.
adjust the bias setting resistor and increase Vre to about 5mV.
Check the temperature of both the heatsink and the output devices after about 30minutes.
If cool then adjust to Vre=10mV. Again check the temperatures.
If warm leave like that and listen for a couple of weeks.
If still cool adjust in 5mV increments to 20mVre. Leave alone and listen for a couple of weeks.
Be very careful the amp is not going to overheat before increasing Vre>20mVdc.
Note that the protection locus of the current limiter has been altered in this tech info. Do all these modifications as a set.
the part you need is on the last page.
It shows a two stage EF output.
The emitter resistors are 0r22.
Optimal output bias will be ~20 to 25mV across each emitter resistor.
Short the input of the amp (this should have the same effect as turning volume control to minimum).
Remove the speaker cables.
Measure the DC voltage across Re. It should be mV not volts and not zero.
adjust the bias setting resistor and increase Vre to about 5mV.
Check the temperature of both the heatsink and the output devices after about 30minutes.
If cool then adjust to Vre=10mV. Again check the temperatures.
If warm leave like that and listen for a couple of weeks.
If still cool adjust in 5mV increments to 20mVre. Leave alone and listen for a couple of weeks.
Be very careful the amp is not going to overheat before increasing Vre>20mVdc.
Note that the protection locus of the current limiter has been altered in this tech info. Do all these modifications as a set.
Last edited:
Hi
What Andrew said is good - 20 to 25 mV across each emitter resistor.
I wouldn't wait a couple of weeks, but the idea to do it in small increments and allow time for the temperature to settle is good.
Now the important bit - where to find them:
The pic below is the right-hand side of page 17.
The resistors are circled in red, but it looks like it will be easier to measure where the green markings are. It looks like connections there to little separate circuit boards? Maybe easier to find those than the resistors.
e.g. If you adjust for 40mV between (A) and (B), that will give you 20mV across each of the top two resistors.
Similarly, 40mV between (C) and (D) means there's 20mV across each of the bottom two resistors.
The trimpots to adjust are R329 and R429. The picture on page 3 of the manual (page 2 of the PDF) shows where to find them.
Do it super-carefully though. Setting the current too high by mistake might blow the output transistors.
Good luck
Cheers - Godfrey
edit: Sorry, missed a couple of posts while I was figuring this out.
It should be about 20mV across one resistor (e.g. top to center pin) or about 40mV from top to bottom.
What Andrew said is good - 20 to 25 mV across each emitter resistor.
I wouldn't wait a couple of weeks, but the idea to do it in small increments and allow time for the temperature to settle is good.
Now the important bit - where to find them:
The pic below is the right-hand side of page 17.
The resistors are circled in red, but it looks like it will be easier to measure where the green markings are. It looks like connections there to little separate circuit boards? Maybe easier to find those than the resistors.
e.g. If you adjust for 40mV between (A) and (B), that will give you 20mV across each of the top two resistors.
Similarly, 40mV between (C) and (D) means there's 20mV across each of the bottom two resistors.
The trimpots to adjust are R329 and R429. The picture on page 3 of the manual (page 2 of the PDF) shows where to find them.
Do it super-carefully though. Setting the current too high by mistake might blow the output transistors.
Good luck
Cheers - Godfrey
edit: Sorry, missed a couple of posts while I was figuring this out.
It should be about 20mV across one resistor (e.g. top to center pin) or about 40mV from top to bottom.
Attachments
Last edited:
this is fantastic info guys, thanks for all the help
luckily the resistors and the trimpots are on a seperate PCB on the top of the motherboard, measuring should be fun as the leads from the resistor assembly are very short,almost flush with the board and have insulators on each of them, so i'll desolder them and add slightly longer pins on so i can attach the multimeter leads
ill let you know how it goes on, and more importantly how it sounds!
thanks very much
luckily the resistors and the trimpots are on a seperate PCB on the top of the motherboard, measuring should be fun as the leads from the resistor assembly are very short,almost flush with the board and have insulators on each of them, so i'll desolder them and add slightly longer pins on so i can attach the multimeter leads
ill let you know how it goes on, and more importantly how it sounds!
thanks very much
It wouldn't necessarily give better sound, you can run into a phenomenon called gm doubling where basically there becomes a kink in the transfer characteristic. Also you run the risk of overheating - amps have barely big enough heatsinks to cope with standard bias let alone extra bias. You reduce headroom as well.
Increasing the bias current increases the heat output, which may or may not damage the amplifier.
The sound quality issue is interesting, though.
I remember John Linsley Hood, for example, recommending higher than "ideal" idling currents in class AB amps. While it results in higher distortion at high output powers, he considered this a fair trade for low distortion class A operation at low output levels.
This seems like a reasonable compromise given that most listening involves low average power, with only occasional peaks of high power (except where teenagers and/or alcohol is involved, in which case sound quality is not a primary concern anyway).
No, BJTs. This was in 1970. Here's the article: The Class-A Amplifier Site - JLH Class-AB Amplifier
The bias issue is discussed in the "Letters to the Editor" section.
Cheers - Godfrey
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.