Measuring Rbulk of transistors

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: Joe., lisa, luketeaford, Kent

User avatar
devinw1
Super Deluxe Wiggler
Posts: 1375
Joined: Tue Aug 07, 2018 11:20 am
Location: Portland, OR
Contact:

Re: Measuring Rbulk of transistors

Post by devinw1 » Fri Feb 14, 2020 11:26 am

I measured the DMMT3906 with the MAT02 setup yesterday and got 0.58 Ohms. Caveat for this one is I had to halve the current from 9mA to 4.5mA. At 9mA (the current listed in the MAT02 datasheet (9Vpp thru a 1k)) the output waveform was clipped on the top.

The THAT320 is another story. For some reason I can't get anything out of the circuit with this chip. I've tried 2 different chips that should be just fine and the instrumentation amp (INA121 running at +/-12V) just goes to the rail. No ramp wave out like all the others. I triple checked the wiring and also tried lower gains of the ouput amp. It seems like the differential voltage is way too large so the amp just goes to the rail. I dunno maybe I have the terrible luck that half of the transistors are bad on both these chips. Seems very unlikely. I've never had an issue with THAT320s before. :despair: :despair:

User avatar
guest
Super Deluxe Wiggler
Posts: 4186
Joined: Mon Aug 19, 2013 11:49 am

Re: Measuring Rbulk of transistors

Post by guest » Fri Feb 14, 2020 1:14 pm

thats odd. with a gain of 100, it should be 10V at the output, max. can you probe the input to make sure they are swinging within range of each other? is the REF pin on the instrumentation amp at 0V or lower?

also, just noticed that im running my Rbe test at 100uA->1mA, and the MAT02 is 1mA->10mA.
openmusiclabs.com

User avatar
devinw1
Super Deluxe Wiggler
Posts: 1375
Joined: Tue Aug 07, 2018 11:20 am
Location: Portland, OR
Contact:

Re: Measuring Rbulk of transistors

Post by devinw1 » Fri Feb 14, 2020 1:25 pm

Yeah, I'll check the inputs when I get a chance. Ref is tied to 0V.

On the current, did we not conclude in earlier discussions that any of these rBE tests should be run at higher current to be useful though, due to the nonlinearities at lower base currents?

User avatar
guest
Super Deluxe Wiggler
Posts: 4186
Joined: Mon Aug 19, 2013 11:49 am

Re: Measuring Rbulk of transistors

Post by guest » Fri Feb 14, 2020 2:09 pm

thats only for the schmitz style test, because Vbe changes are larger than Rbe changes at low currents. but, with this double transistor method, the Vbe changes can be cancelled out, allowing you to zoom in on the Rbe changes. the signal is noisier, but somewhat readable down at that level (to within 10%). most of the noise in my setup was 60Hz hum, as i wasnt too careful with traces and shielding. in theory, the mat02 method should be even better due to the diffamp.
openmusiclabs.com

User avatar
devinw1
Super Deluxe Wiggler
Posts: 1375
Joined: Tue Aug 07, 2018 11:20 am
Location: Portland, OR
Contact:

Re: Measuring Rbulk of transistors

Post by devinw1 » Fri Feb 14, 2020 3:39 pm

Aha, cool. That makes sense. Thanks for the explanation.

I can easily change the current down a factor of 10 with the MAT02 setup and then change the diff amp gain to 1000 so I can try both. Signal was very clean at gain 100 so might be fine at 1000 too. Usable enough anyway.

User avatar
guest
Super Deluxe Wiggler
Posts: 4186
Joined: Mon Aug 19, 2013 11:49 am

Re: Measuring Rbulk of transistors

Post by guest » Fri Feb 14, 2020 7:45 pm

i just measured the 2164:

SSM2164 - 1.6ohms
SSI2164 - 0.7ohms

i tried to measure the LM13700, but it is behaving very strange. i think the current mirrors inside are not linear enough.
openmusiclabs.com

User avatar
devinw1
Super Deluxe Wiggler
Posts: 1375
Joined: Tue Aug 07, 2018 11:20 am
Location: Portland, OR
Contact:

Re: Measuring Rbulk of transistors

Post by devinw1 » Sat Feb 15, 2020 4:02 pm

OK, finally got some numbers out of the THAT320! I don't know what was going on before. Below are some pairs measured with the MAT02 circuit, only with 2 test current windows. One maxes at 4.5mA (half the MAT02 circuit recommendation) and the other at 900uA ("real world scenerio).

In general you get a little higher rBE measurements with the lower current, and the THAT320 agrees with the datasheet very well (I think they say typical 1, max 2). :sb:

The data:
Screen Shot 2020-02-15 at 12.59.12 PM.png
You do not have the required permissions to view the files attached to this post.

User avatar
Trippin
Learning to Wiggle
Posts: 9
Joined: Tue Dec 24, 2019 3:43 pm

Re: Measuring Rbulk of transistors

Post by Trippin » Sun Feb 16, 2020 6:05 pm

This may seem obvious and my understanding is not great but the results show that the THAT320 is still king but if you wanna be cheap too lazy to hand match, the dmmt3906 is the way to go. Is this right or is my understanding weak?
Kinda sus

User avatar
guest
Super Deluxe Wiggler
Posts: 4186
Joined: Mon Aug 19, 2013 11:49 am

Re: Measuring Rbulk of transistors

Post by guest » Sun Feb 16, 2020 6:17 pm

its the other way around. a higher value is worse in this case. so the NST4501 is the best on this list.
openmusiclabs.com

User avatar
devinw1
Super Deluxe Wiggler
Posts: 1375
Joined: Tue Aug 07, 2018 11:20 am
Location: Portland, OR
Contact:

Re: Measuring Rbulk of transistors

Post by devinw1 » Mon Feb 17, 2020 1:12 pm

Right, from the rBE standpoint (lower is better), and my experience in actual testing, several of the surface mount PNP pairs work very well for exponential converters. The NST4501 is great, and also the PMBT4403 is very good but it is not technically marketed as a matched pair even though the small sample of units I have all had great matching for Vbe and HFE. There are probably other part numbers too, but pretty much all of them will be the TSOP-6 package which is sorta tiny :D.

In the PNP world, the THAT320 is still a very good chip, and is in rare DIP form which is cool for through hole but it is a quad, so using only half of it is pretty wasteful since it's not cheap (like $7). Most of these SMT pairs are less than a dollar.

loki
Common Wiggler
Posts: 91
Joined: Fri Apr 25, 2014 4:20 pm

Re: Measuring Rbulk of transistors

Post by loki » Mon Feb 17, 2020 3:48 pm

A couple of times in the past I have posted suggesting the use of the MMDT4401 (NPN) and MMDT4403 (PNP) as exponential pairs. The late Marshall Leach recommended the 2N4401 and the 2N4403 as the secret sauce in many high end audio designs. Because of their higher current rating they have a bigger base-emitter area which reduces Rbe and Rbb. The lower Rbb results in lower noise in certain amplifier circuits. It's nice to see the lower Rbe confirmed.

User avatar
neil.johnson
Super Deluxe Wiggler
Posts: 1336
Joined: Thu Apr 08, 2010 3:51 pm
Location: Cambridge, UK
Contact:

Re: Measuring Rbulk of transistors

Post by neil.johnson » Tue Feb 18, 2020 5:05 pm

guest wrote:
Fri Feb 14, 2020 7:45 pm
i just measured the 2164:

SSM2164 - 1.6ohms
SSI2164 - 0.7ohms
That's not surprising with the larger transistors in the SSI2164 giving lower resistance and thus better overall log conformance.

Neil
Random ramblings :: http://www.njohnson.co.uk

User avatar
Trippin
Learning to Wiggle
Posts: 9
Joined: Tue Dec 24, 2019 3:43 pm

Re: Measuring Rbulk of transistors

Post by Trippin » Tue Feb 18, 2020 6:56 pm

Are you planning to do NPN pairs? What NPN complement to the NST4501 do you suggest?
Kinda sus

User avatar
guest
Super Deluxe Wiggler
Posts: 4186
Joined: Mon Aug 19, 2013 11:49 am

Re: Measuring Rbulk of transistors

Post by guest » Tue Feb 18, 2020 8:22 pm

neil.johnson wrote:
Tue Feb 18, 2020 5:05 pm
guest wrote:
Fri Feb 14, 2020 7:45 pm
i just measured the 2164:

SSM2164 - 1.6ohms
SSI2164 - 0.7ohms
That's not surprising with the larger transistors in the SSI2164 giving lower resistance and thus better overall log conformance.

Neil
neil, do you know why they increased the control port impedance to 1k+9k from 500ohm+4.5k? it seems like that would add noise and reduce log confromance. and from what ive heard, its kept at 500ohm+4.5k for the new SSI2162, and 250ohm+2.25k for the SSI2161.
openmusiclabs.com

User avatar
neil.johnson
Super Deluxe Wiggler
Posts: 1336
Joined: Thu Apr 08, 2010 3:51 pm
Location: Cambridge, UK
Contact:

Re: Measuring Rbulk of transistors

Post by neil.johnson » Wed Feb 19, 2020 11:46 am

guest wrote:
Tue Feb 18, 2020 8:22 pm
neil, do you know why they increased the control port impedance to 1k+9k from 500ohm+4.5k? it seems like that would add noise and reduce log confromance. and from what ive heard, its kept at 500ohm+4.5k for the new SSI2162, and 250ohm+2.25k for the SSI2161.
The SSI2162 and SSI2161 are simply internally-paralleled SSI2164 die, so you can work out why the input resistance of the control pins is half or quarter that of the SSI2164.

Neil
Random ramblings :: http://www.njohnson.co.uk

Post Reply

Return to “Music Tech DIY”