Silly question about making due with trimpot values

[Kevin Mitchell]

I'm working on my first phaser and as always I try to make due with what I have with these quick builds.

I don't have a 250k trimmer and the best I can do is 100k. I've seen a schematic that used a 25k trimmer with additional resistors on the outside and lowering the 1M on the wiper I plan to tinker with a similar setup. Schematic link here

I can't yet wrap my head around how to go about using a lower trimpot and how I would know if I'm close to "the sweet spot". Any help would be very much appreciated.

Thanks guys!

[GibsonGM]

It's not a silly question at all, Kevin     Trouble is, not knowing where the sweet spot is, like you said.

You CAN add 100k to 150k in series with your trimmer, to "make it a near 250k", which is variable from 250kish to 100k.    If the sweet spot is as 175k, boy, yer golden!   Depending on the circuit, if the sweet spot was 75K, you'd be golden with just the trimmer.  This won't work in some instances, tho, if the RATIO between the 'halves' of the trim pot is important (does that make sense?)

Can you post the specific schematic you want to build from?  That would be much more helpful in how you 'get out of this'.     Do you have an actual 250k pot available?  Sometimes you can tack one in, just to see where that sweet spot it...it can give you info, but not really if you need 4 of them or something


Please show us where you need to mod.

[Kevin Mitchell]

I've redrawn the schematic to help me breadboard it all but I'm using this schematic slightly modified. Schematic here

I believe I have 4 matching 2n5457 (4 of 10 matched close, luck?  ) And the pinout should be the same as a 2n5952 (DSG). Also I'm using 47n caps instead of the 50n and LM358 opamps.

Mods? No mods yet. I'll likely add high and low controls for the depth, add a control for subtle external biasing and also a couple extra stages once I have enough matched fets. Oh and an LFO LED. These are standard mods really. But I'm not there yet.

[GibsonGM]

Boo yeah, there ya go Kevin.  For the bias control, if it were I - I would temporarily tack in a 250k POT and get it adjusted, then measure the pot.  See what the resistances to both ends from wiper are.    Subtract those values plus some more, use fixed R's at either end, and fine adjust with your pot.  For those who feel lucky, they might just put the fixed resistors in and be done (omit trimmer).  I like the option to tweak, however.

If you simply guessed, and put a 47k resistor on each end, you might luck out...or, you might find the nice spot is at 33k and you can't get there, LOL.

If you find this is a bit more work than you'd like, just order the right trimmer!!    It stinks to wait for the shipment to come, I know... 


By 'mods', I just meant where you needed to vary from the schematic's specs with this trimmer.

[Kevin Mitchell]

I believe I could snag a 250k rev log pot from a scrapped practice amp. Not sure if it being type C would compromise anything. Maybe I have a 500kB pot I could half...

Thank you very much for your explanation.

[Kevin Mitchell]

Now that I think about it, it shouldn't matter what type of taper it is... There's my brainfart of the day.

[GibsonGM]

YW, man, welcome to the addiction, lol!   Yes, the C taper will work fine for ya!  Just turn slow.

I suspect the designer wanted the 250k there for a reason, or they'd have made it 100k.  You never know, they may have had 100,000 extra 250k trimmers lying around!   

I'd like to know what you come up with when you measure it....

[duck_arse]

if it was me, I'd get a 100k resistor and the 100k trimm. take a punt, put the resistor to the low side, ground, and the pot to the high side, whatever that voltage is. sweep the pot, even though it only provides half the full sweep of the reference. find yr sweets? if not, swap trimm and resistor places, try again. find it this time?

because you're sweeping 100% of the voltage with the 250k and its 270 degrees rotation, and only 50% with the 100k+100k pot and the same 270 degrees, you have increased your adjustabillity/sensitivity/granularity/I don't know the right term.

same applies when you find the rough sweets, you can measure the resistance from the centre to each end of the string, and then put maybe a 50k trimm, with resistors either side of your measured to total 250k. and you can even panel mount that pot, with those fixed R's as stoppers, to sweep the different phaser tones that come with different bias points.

[GibsonGM]

Good one, Duck! 

[PRR]

> I suspect the designer wanted the 250k there for a reason, or they'd have made it 100k.  You never know, they may have had 100,000 extra 250k trimmers lying around!

I'm sure a 250K fell to hand when needed.

What are we biasing? Four JFET gates. What is their current? Dang near zero. We could feed them from a 100 MEG source. We can certainly be near 1Meg, because the trimmer is padded-out with 1Meg as part of the wobble-divider with the 3.3Meg.

What is our other limit? Probably battery life, waste power, heat. A 1 Ohm pot would pull maybe 9 Amps (well, 3 Amps because it shows a 3V reference), which is a lot of Power, and FAR more power than we need to tickle some JFETs.

25K makes sense because such a low value will show 1% drop when poked with a 1 Meg voltmeter, yet does not suck big power.

250K makes sense because we don't need 1% meter-load accuracy, and for some dudes 250K trimmers are like candy.

10K 50K 100K 500K all much the same. At 1Meg you might think about the 0-250K wiper impedance against the 1Meg divider resistor, and wonder if the "1Meg resistor" might like to be 800K or so. Probably not critical; if it is (because the JFETs' Vp is not what this designer intended) then this too could be a trim (argh!!).

[Kevin Mitchell]

Just a little report for the morning; I've got it running on the breadboard this weekend using the scrapped 250k pot. The sweet spot is at about 3/4 rotation on the rev log. So there's a chance I could get the spot with just the 100k trim by lowering the 1M resistor as PRR was saying. I'll find out later.

I'll also give you guys some readings for reference later today. The reason a 250k trim was used was to accommodate the variation in whatever matched fets you are using since every matched set doesn't measure the same. 250k may be more than enough as I suspected -or at least for my set.

Report on the sound; it seems to hang on one side of the phasing a bit more. With the rate up it sounds very much like the univibe from Hendrix's Rainbow Bridge performances. I dig it allot but it's not what I was expecting. It really reminds me of my 6 stage Easyvibe. Perhaps a breadboard error (LFO) or just the results of the FETs not being match close enough so I'll check that out as well.

PRR, thanks for the intel. Food for thought makes us stronger...

A question; any idea what waveform should be coming off the LFO? I'll scope that as well.

[PRR]

> what waveform should be coming off the LFO?

There are many ways.

What waveform are you seeing?

Sine is classic.

Triangle is a lot easier to generate.

"Hang" is likely to be too-big wave.

BTW: "make do".

[Kevin Mitchell]

Ehh I can't get a read with my setup. But with an LED on the ouput of the LFO opamp I am seeing a square wave.

After dialing in the 250k more and checking out some stock and DIY phase 90s on youtube I think it's as good as it's going to get as far as a phase 90 clone goes.

I'm having a little trouble understanding how it works due to my lack of knowledge in electronics. If I were to guess I'd say the output of the LFO is pushing the FETs to shift (as a resistor) and it's the FETs themselves that produce the sine wave phasing given they are dialed in well enough.

I'm reading 3.57 volts on the wiper and 4.07 on lug 3 of the 250k. V+ is 9.32 (all measurements move + or - .01 volts with the lfo)

Here's a clip. Got carried away for a sec...

BTW: "make do".

Lol Thanks.

[PRR]

> trouble understanding how it works ...to guess I'd say the output of the LFO is pushing the FETs to shift (as a resistor) and it's the FETs themselves that produce the sine wave phasing

Explain why there are 0.05 caps in there.

FWIW: the JFETs won't "phase shift" much below 10MHz. To us, they are resistors, voltage-variable. Resistors alone don't phase-shift.

[Kevin Mitchell]

Since the Phase 90 and the Easyvibe schematics are very similar I can't help but to think of the FETs as variable resistors in this case (I mean LDRs) and the source (2nd pin of FET) is like light to the LDR. But since the output of the LFO seems to be a square wave I was wondering what was causing the sound to be more of a sweep as a phaser does so I thought it was the FETs themselves. So I'm wrong?

The sweep doesn't seem to come from the LFO but is triggered by the LFO is what I mean.

Explain why there are 0.05 caps in there.

If I've studied correctly the capacitors rely on the variable resistor (FETs in this case) to sway it's capacitance to shift the signal in the phase stage.

As I've said I don't have much knowledge in electronics but I'm learning along the way.

[PRR]

> the output of the LFO seems to be a square wave

The output of the *chip* IS a square wave.

However the Gates are driven from the top of C10 15uFd(?). That fat cap obviously can't hop up-and-down real fast. Especially with 5K to 505K connecting it to the chip out pin. This R-C network will tend toward an exponential, both ways since the chip is square-waving. But if the swing is small, it will be nearly triangular wave (you can't see the curve in the exponential until it goes on a ways).

The classic trem will use a Sine. But these are hard to make cheap, especially slow. And the ear hears "pitch" different from amplitude. A tri-wave sweep makes sense, and is real cheap, even when slow.