match transistor for xc-phase (banzai music, retroamplis)

[NickC]

some time ago I bought a set of 2n5952 already matched (banzai music), but being curious I tried to build a circuit to test them, I used this http://www.geofex.com/article_folders/fetmatch/fetmatch.htm
now I would like to know, I have to place the multimeter on the gate and on the 3 pin of the TL072 or the gate and source are fine too?
I noticed that I get the same result.

I noticed (if I did everything correctly) that the transistors are matched but in pairs of 2 and not in 4
0,26
0,26
1,88
1,88

wtf.

 

[mkstewartesq]

Following. I just bought a matched set of the same transistors from Banzai for an XC Phase build that, once I installed them, is not phasing properly. This led me to have concerns this perhaps they were matched (although I would’ve expected all four to be matched, not two and two) but not necessarily matched to phaser specs (Banzai, unlike other sellers of matched quads, doesn’t specifically state that they are good for a phaser in the product description). So I am lucky to be having a friend here assist me in acquiring a matched set that has been matched to be appropriate to a phaser, just in case the Banzai are not.

Mike

 

[NickC]

ok , i also used this circuit to match the transistors, i can confirm that the ones sold by banzai are not matched to 4 but in pairs of 2, instead i ordered a set from retroamplis and they are perfectly matched, in fact now the phaser sounds wonderful, with the banzai set sounded decent but now that's another story.

2.30 vp 0.60 idss all 4 transistor!

 

[benny_profane]

is not phasing properly

There’s only a narrow part of the bias trimmer that will work. Try dialing it in more slowly.

 

[owlexifry]

i bought x2 matched quads 2N592 from banzai.
used the same matching circuit @NickC listed above.
didn’t really find a huge variation between them all, picked the closest x4 of the x8, and worked just fine in my vero phase 90 build.

 

[mkstewartesq]

The only anecdotal evidence I can give is that which I reported on in more detail in my build thread - with the matched set form Banzai in my XC Phase, my phasing in all modes was somewhat asymmetrical, and 2 stage/"45" mode was completely unusable due to distortion that could not be dialed out. Once I replaced them with a set matched by a member here and kindly provided to me, both issues disappeared.

Mike

 

[mkstewartesq]

There’s only a narrow part of the bias trimmer that will work. Try dialing it in more slowly.

Just saw your comment - see above. The issue was definitely with the Banzai matched quad.

Mike

 

[owlexifry]

Just saw your comment - see above. The issue was definitely with the Banzai matched quad.

Mike

did you measure your quad to confirm how close they are?

 

[mkstewartesq]

did you measure your quad to confirm how close they are?

No, I didn’t, at least not using the RG Keene method (I need to buy the Rullywow testing PCB because I don’t have the knowledge yet to convert the schematic into a testing circuit myself yet.).

I did check the voltages on the Banzai set when it was phasing and was getting about the same reading on the gate for each (somewhere around 2.1 V to 2.3 V); the drain on two of them was about 4.5 V and on the other two about 5.5 V – but this was when it was in circuit so I have no idea if that differential was intended or not or whether it really says anything about how well they were matched. As far as the replacements that I got from Fig here, I just trusted that they were matched more properly, especially since Fig matched them himself, they were basically plug and play and fixed the issues I was having. Interestingly, when phasing, the gate voltage on each of the new set were closer to 1.6–1.8 V when the supposed sweet spot for phasing is between 2 V and 2.5 V - but they work just great.

Mike

 

[owlexifry]

No, I didn’t, at least not using the RG Keene method (I need to buy the Rullywow testing PCB because I don’t have the knowledge yet to convert the schematic into a testing circuit myself yet.).

I did check the voltages on the Banzai set when it was phasing and was getting about the same reading on the gate for each (somewhere around 2.1 V to 2.3 V); the drain on two of them was about 4.5 V and on the other two about 5.5 V – but this was when it was in circuit so I have no idea if that differential was intended or not or whether it really says anything about how well they were matched. As far as the replacements that I got from Fig here, I just trusted that they were matched more properly, especially since Fig matched them himself, they were basically plug and play and fixed the issues I was having. Interestingly, when phasing, the gate voltage on each of the new set were closer to 1.6–1.8 V when the supposed sweet spot for phasing is between 2 V and 2.5 V - but they work just great.

Mike

having ordered a matched set/quad, you would certainly hope that they do indeed 'match', but i guess when you're troubleshooting, it's important to confirm for yourself, as many variables/ unknowns as possible.

even though it's sort of irrelevant now, you could easily build this. there aren't even any track/trace cuts to worry about.

using this, you could actually confirm how much variation there is in your 'matched' quad from banzai.

(and possibly even use it as evidence to request a refund / replacement quad, since it appears you didn't get what you paid for)

 

[mkstewartesq]

having ordered a matched set/quad, you would certainly hope that they do indeed 'match', but i guess when you're troubleshooting, it's important to confirm for yourself, as many variables/ unknowns as possible.

even though it's sort of irrelevant now, you could easily build this. there aren't even any track/trace cuts to worry about.

using this, you could actually confirm how much variation there is in your 'matched' quad from banzai.

(and possibly even use it as evidence to request a refund / replacement quad, since it appears you didn't get what you paid for)

Actually, that’s very helpful, thank you. I might still have some other situations where I need to check the vgs(off) on a transistor, which I understand this does. Having never worked with strip board, the fact that there are no cuts to worry about means that I can basically just buy a piece of stock strip board and go to assembling the components on it as per the diagram, correct?

On a quite possibly unrelated note, I have a Proposition board for that requires a JFET with fairly specific vgs(off) and rds(on) characteristics. Is anyone aware of any method to test for both of these values, short of buying a Peak DCA Pro 75? While that is certainly a nice piece of kit, I honestly don’t see myself making such frequent use of its more advanced features that I can justify the $200 price tag for myself.

Thank you again,
Mike