Matching Mosfets

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Gate voltage on the 2SJ50/2SK135 Hitachi

The gate has a voltage value where the fet "opens". Is it nessesary to messure this voltage value to state whether the two fets is considered to be a good match? If so ...how do we messure this ? by a 10K multiturn pot from the 22Vdc supply to the gate?

How do we messure the gate voltage ( when the fets opens) on the Hitachi's 2SJ50/2SK135?

Anyone who has the circuit?
 
Nelson wrote an article about matching long time ago,

http://www.passdiy.com/pdf/matching.pdf

Exact circuit simply uses a potentiometer between source and supply voltage with the wiper attached to its gate.

Take care to turn the potentiometer to the source when changing the fets as otherwise bad things can happen ;)

Oh and last, you certainly don't need to match at 1.125 A as Eric does, that's a pretty special application. Check your amp design and see at which bias the outputs run.

Otherwise 100-200mA should do fine.

For your 10 pcs that's really sufficient.

Have fun, Hannes
 
Try to get the point!

Nelson wrote an article about matching long time ago,

http://www.passdiy.com/pdf/matching.pdf

Exact circuit simply uses a potentiometer between source and supply voltage with the wiper attached to its gate.

Take care to turn the potentiometer to the source when changing the fets as otherwise bad things can happen ;)

Oh and last, you certainly don't need to match at 1.125 A as Eric does, that's a pretty special application. Check your amp design and see at which bias the outputs run.

Otherwise 100-200mA should do fine.

For your 10 pcs that's really sufficient.

Have fun, Hannes

Ok - I try to use the curcuit by Nelson Pass from 1993, and the potentiometer you suggest...and yes...remember to turn to source by change of devices.

About the design the amplifier should run with approx. 100mA per pair of Hitachi's.

Thanks Hannes.

rgds.
 
Last edited:
it's not the Gate voltage you measure.
It's the Vgs that should be measured.
Now you have to decide @ what Id you want the Vgs while Vds is set to ?
and Tj=? and Tc=?

Repeat from the first message:
"Hi , as being new in electronics I need a simple circuit to match a double pair of 2SJ50 / 2SK135 Hitachi mosfets for a 100watt amplifier."

To Mr. Andrew T. by the Scottish Borders, If one knew how to match power mosfets and understood the function Id /Vgs while Vds is set to? and Tj=? and Tc=? , one might just have matched the fets.
 
Last edited:
Hi Godmund

I have been asking the same question and started a tread
sorry to see you incoured in the Mr T plague

My opinion on this are quite clear just use your ignore list save loads of space when searcing.

Any way Frags bless him and the one like him came up with a couple of real good suggestions.

I am working on it and at the moment I got preaty good results in about a couple of minutes (the temperature rise keep changing the results as Vgs is temperature dependent)

As soon as I tidyed up things I will post .

For now look at this
http://www.diyaudio.com/forums/attac...1&d=1291953183

and post 47 on my tread

At work at presen so sorry for limited response
more to folow

Al
 
Work with Hitachi's.

Thanks Al,

At work this morning.

Happy to hear that I am not the only one in contact with the Scottish Borders..-)))

Working as eyespecialist and soon close to retirement, electronics is not my field. Without electronic education I have made 6 different amplifiers with bipolar transistors...they all plays music beautifully. The old Hitachi's has a "sort of a warm sound", I think.
If you could help me on this one (matching), I could test your eyes??:)

Just fun...look forward to read your comments later.

Kind regards
Gudmund
 
If you could help me on this one (matching), I could test your eyes??

;)

I would help you right now if I had the means for the old TO-3 packages. I know they work well and have great thermal properties, but I always stayed away from these metal cans.

Sorry, however a free eye check is really a nice offer ;)

Don't worry you get that done!

Have fun, Hannes
 
it's not the Gate voltage you measure.
It's the Vgs that should be measured.
Now you have to decide @ what Id you want the Vgs while Vds is set to ?
and Tj=? and Tc=?

Repeat from the first message:
"Hi , as being new in electronics I need a simple circuit to match a double pair of 2SJ50 / 2SK135 Hitachi mosfets for a 100watt amplifier."

To Mr. Andrew T. by the Scottish Borders, If one knew how to match power mosfets and understood the function Id /Vgs while Vds is set to? and Tj=? and Tc=? , one might just have matched the fets.
there are just three sentences in my reply.
the first tells you do not try to measure Gate voltage.
the second tells you to measure Vgs.
Pause.
Why is gate voltage wrong? What is Vgs? How do I find out? Who can I ask? Is there a way to find this information myself?
As a newcomer all the jargon will be new/unknown to you.

the third sentence defines what must be done to measure the Vgs. It has even more jargon. Help me.
Yes we will all come in and describe what each of us think is the simplest way to lead a newcomer through this topic and all the others you are going to bump into.

I'd suggest you go back to Pause and think about those questions that should already be whirring around in your head, rather than saying: ugh, I don't (want to) understand.
 
I'd suggest you go back to Pause and think about those questions that should already be whirring around in your head, rather than saying: ugh, I don't (want to) understand.

Impressed. After a few messages you have turned the simple issue of matching some fets to know what one want's to understand.

It's not the few lines in the received reply that could disturb a nights sleep, it's the idea and attitude.

You ask me: I have got keratoconus on both eyes and I want to be able to see again.

My answer: You have to deside : Boston Xo material, scleral or aspherical, if aspherical deside 28 axes or 4 axes, please state approx. AEL edgelift before you deside. Please confirm POTF thickness, amorph, waves or??

alternative: Corneal transplantation, please state approx. CT, Iris DM? and....more?

Would you say something like: "Please speak english to me"?

Your suggestion: Yes.. lets take the third line: "Pause".

Best regards
Gudmund
 
Last edited:
is the setup for IR fets the same as for Hitachi's?

Hi Godmund


I am working on it and at the moment I got preaty good results in about a couple of minutes (the temperature rise keep changing the results as Vgs is temperature dependent)

As soon as I tidyed up things I will post .

For now look at this
http://www.diyaudio.com/forums/attac...1&d=1291953183

and post 47 on my tread

At work at presen so sorry for limited response
more to folow

Al

Hi Al,

Have looked at the matching setup from your post 47. Is the amount of mA messured enough to state: matched or device too far from each other?

Could I use the same setup for Hitachi's ( called lateral? right?) as for the IRF types ? The setup need to be changed for NPN and PNP types?

Hope you got freetime for this!

Rgds
Gudmund
 
Hi the different current indicate mismatch
Not sure about laterals I suppose just different voltage to gate but better asks any that used them.

NpN PnP you need to swap voltages around. See water example they flow a different way.

Grollins post 14 is quite right try to match at similar to working conditions that preaty much answer all the voltages and current questions you may have

DR. Gone post 16

Quite elegant solution similar to what Frags posted on my tread

This is similar to what I will be using +
A) I am trying to keep the temperature constant so good thermal coupling between he 2 small heatsink and variable speed fan.

2 the resistors to the Drain need to be matched unless you keep one on the heat sink as temperature equalisation and do the swap on the other

If you have similar readings between the ones you swap they be matched

EG Right hand as fixed and tests on the left you may have –100 mV on one and –105mV on the other those 2 will be preaty close.

While you there have a play with the thermal coupling E.g. see what happens if you use grease or whatnot or if you leave the clamping screw a bit loose.

You be surprised how important decent thermal coupling is.

Got reading al over the place just by pushing on the case of one with a screwdriver.

Papa write up is quite good, one thing it does not mention is that when you match power dudas the 10 ohm resistor should be a 25 W minimum.

PS last but probably the most inportant bit buy a few more than what you need and maybe a few cheap ones whyle you crack this
Best way to get it right is to blow a few.
 
Hi the different current indicate mismatch
Not sure about laterals I suppose just different voltage to gate but better asks any that used them.

NpN PnP you need to swap voltages around. See water example they flow a different way.

Grollins post 14 is quite right try to match at similar to working conditions that preaty much answer all the voltages and current questions you may have

DR. Gone post 16

Quite elegant solution similar to what Frags posted on my tread

This is similar to what I will be using +
A) I am trying to keep the temperature constant so good thermal coupling between he 2 small heatsink and variable speed fan.

2 the resistors to the Drain need to be matched unless you keep one on the heat sink as temperature equalisation and do the swap on the other

If you have similar readings between the ones you swap they be matched

EG Right hand as fixed and tests on the left you may have –100 mV on one and –105mV on the other those 2 will be preaty close.

While you there have a play with the thermal coupling E.g. see what happens if you use grease or whatnot or if you leave the clamping screw a bit loose.

You be surprised how important decent thermal coupling is.

Got reading al over the place just by pushing on the case of one with a screwdriver.

Papa write up is quite good, one thing it does not mention is that when you match power dudas the 10 ohm resistor should be a 25 W minimum.

PS last but probably the most inportant bit buy a few more than what you need and maybe a few cheap ones whyle you crack this
Best way to get it right is to blow a few.

Thanks for all the info. I made a print this morning and will look for post 14 and 16 later today.

Might have some BUZ900/905 to blow. Do you connect your hand with a piece of cobberwire to the radiator ( ground) before you touch the device ...the gate is very sensitive on powerfets?

Thanks for your help

Gudmund
 
Do you connect your hand with a piece of cobberwire to the radiator ( ground) before you touch the device ...the gate is very sensitive on powerfets?
copper wire to your worksurface.

Big mosFETs have a lot of gate capacitance. As a result the voltage does not change much when some static charges up the gate. = few failures.
jFETs seem to be tolerant of static on the gate = no failures.
Small mosFETs and mosFET input chips without input protection are your most likely candidates for careful handling.
 
static charge.

copper wire to your worksurface.

Big mosFETs have a lot of gate capacitance. As a result the voltage does not change much when some static charges up the gate. = few failures.
jFETs seem to be tolerant of static on the gate = no failures.
Small mosFETs and mosFET input chips without input protection are your most likely candidates for careful handling.

Ok - some of the information about static charge on the net told us that the mosfetgates could be sensitive. What you tells us is that the gate capacitance on big fets takes most of the electrical spikes, like a normal capacitor accross a connector?
That makes sence to me.

Rgds.
 
You guys are worrying too much, usual precautions when handling parts are plenty and more than sufficient. No need for copper wristbands or ESD-safe work places.

I never had a shot mosfet and I matched hundreds (all types, small signal and the big guys, N & P channel). The guys take 20V and that's quite some.

EDIT: same setup for verticals and laterals, for P-types swap + and - voltages (supply and ground) in comparison to N.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.