Matching IRF610's

[TomT]

Here's a simple method I use for matching IRF610's used in the balzen preamp.
The link is: http://www.geocities.com/zipper1000000/schematic/tl431matching.jpg
Hopefully the tl431 isn't to exotic. The data sheets can be found on the web at on-semi, TI and Zetex.
The tl431 part has always been a favorite part of mine over the years.

Thanks
Tom

 

[rtirion]

Sorry to say the link isn't working
Reply I get is "This page is not available"

Regards

 

[AudioFreak]

try this.........

goto http://www.geocities.com/zipper1000000/schematic/ then click on the link to tl431matching.jpg

 

[TomT]

Thank-you Audio Freak.

Sorry about the link. Just joined DIYaudio.
Trying to these web things working.

Tom

 

[promitheus]

it´s a very good idea. Does this zener always have 2.5 V reference voltage ?

 

[TomT]

The Zetex part numbers are:
ZR431 is the 2.5V
ZR243 is the 1.24V

The On-Semi and T.I. part numbers are:
TL431 is the 2.5V
TLV431 is the 1.24V

See the web sites for the data sheets.
I believe the voltage references are band-gap types.

I have also used the circuit to match npn transistors and n-channel jfets. When using the circuit with n-channel jfets a resistor or diodes must be placed between the source of the jfet and the reference pin to keep the voltage greater than zero between the cathode and reference pins on the 431 part.
Also have used the circuit as a current source in amplifiers.
Tom

 

[promitheus]

This means the ref voltage is always constant. so we can measure the Vgs of the fets. how can someone hook up a irf9610 to measure? And how can someone measure bigger FETS like IRFP240 at 1 amp?

 

[GRollins]

National Semiconductor has similar parts; the LM-336 series.
Bear in mind that the voltage will drift with temperature, but there's a fairly simple circuit you can build with some 1N914s (or other small diodes) to counteract the drift. Also, the adjust pin means just that--the voltage is approximately 2.5V, but may be slightly off due to manufacturing tolerances. You can include a compensation pot in the temperature correction circuit. Check the Nat. Semi. datasheets for details of the compensation circuit.

Grey

 

[TomT]

Sorry the tl431 circuit does not work with p-channel mosfets like the IRF9610. Here is and idea for matching the IRF9610 mosfet. Go to http://www.geocities.com/zipper1000000/schematic/
then click on the link to lm337match.jpg
I have not tried the tl431 circuit at large currents on power mosfets.
Tom

 

[promitheus]

also a very good idea.

 

[rtirion]

I am a little confused.

Is there a significant advantage using a TL431 or LM337 over using a resistor/stabelized powersupply to match FET's.
(The later method as explained in the PASS A75 manual).

I can not see an advantage from a temperature point of view. I noticed significant temp drift when matching my IRFP9610's and IRFP150's for an ALPEH3.

Or am I missing something here?

Regards

 

[TomT]

Perhaps the problem with the A75 matching circuit is the mosfets have a Vdg (voltage drain to gate)of zero? The circuits I showed may have a similiar problem of low Vdg and low power dissipation. Maybe some Vdg is better than none?
Right now I do not have the power supply built for the preamp. But I do have the mosfets.
I was thinking matching at a lower power level could translate into a higher power level match or at least get them close(like the A75 matching circuit). Also without the hassle of heatsinking, soldering, unsoldering parts. Like I said I'm unsure about high curents and high powers. I should present the ideas in a better context. Saying its just an idea, simple or complex, useful or not(like the 377 circuit).
I apologize for not being clear or thinking things out.

 

[Nelson Pass]

Of course we are simply trying to match the
devices, not characterize them particularly.
The Vdg at zero is not a problem as the actual
voltages of interest are the Vds and Vgs which are
identical. Practical experience shows that the
simple technique with a power supply, resistor
and DC voltmeter gives you the results you want
without getting much more elaborate.

 

[rtirion]

TomT, Mr. Pass, thank you both. My confusion is gone. I'll stick with the A75 matching method.
I try to compensate the temp drift with taking the measurements in almost identical conditions
- All FET's measured for exactly the same time. I use 60 seconds per FET (give or take a second or so).
- Prevent any airflow or cooling during the measurement.

I do not know if these precautions are essential, but they worked for me. I have no problems with DC on the output (+/- 20 and 15 mV on each channel) , or temperature drift in my ALEPH 3.

Regards

 

[Nelson Pass]

I only test them for about 5 seconds, but I do make
sure that they all have reached room temperature
before I start.

 

[Asen]

I have succeeded in matching 4 sets of 3 IRFP240s for an Aleph 30 project out of 20 MOSFETs.
Obviously it's far more difficult to match 2 pairs for the input. I've measured 10 devices (IRF9610) at different currents. Here are the results:

15mA 13mA 7.5mA
1. 3.35 3.32 3.25
2. 3.57 3.55 3.54
3. 3.63 3.61 3.60
4. 3.69 3.66 3.59
5. 3.72 3.71 3.71
6. 3.75 3.71 3.61
7. 3.90 3.88 3.87
8. 3.93 3.92 3.93
9. 3.97 3.95 3.93
10. 4.03 4.02 4.98

Do you see 2 pairs (properly matched) between these devices or I should better go buy 10 more?

Thanks!

 

[Nelson Pass]

Obviously you can leave that 4.98 home, but otherwise, you
have lots of matches.

 

[Megabox]

Matching Quantity

Hi all,

I am going to build an Aleph-P 1.7 and have to match 8 x IRF610 and 4 x IRF9610. How many MOFETs should I need to buy if the difference among themself is 0.01v or less? Please advise.

Thanks.

 

[breguetphile]

I'm not sure about IRF610, but I bought some 9610's for an Aleph 3 I'm building and was surprised to find 4 of them matched within 0.02V. Of course I need only need two pairs of two, but if you need four, 10 may be enough, maybe a few more won't hurt .