MXR - Phase 100  [schematic]

Discussion regarding early stompbox technology: 1960-1975 Please keep discussion focused and contribute what info you have...
User avatar
armdnrdy1
Breadboard Brother
Information
Posts: 104
Joined: 07 Oct 2012, 00:11
Has thanked: 10 times
Been thanked: 67 times

Post by armdnrdy1 »

zaphod wrote:That article has a pretty good summary of textbook flanger and phaser theory. R.G. Keen writes some great articles. What's interesting is that he seems to be agree with me on some important points.
It turns out that the human ear is not too sensitive to phase shifts...
The most common commercial phase shifters stop with four stages (such as the MXR Phase 90, Univibe) or six stages (MXR phase 100).
...although the effect is noticeable, it's much better if the notches move up and down in frequency.
So while I might gree that maybe there's some benefit in having some extra fixed phase shift stages, it's really not clear how much of a benfit that really is. Maybe it becomes more useful when feedback is switched on...? :hmmm: Even then, wouldn't it make more sense if different RC values were used in each of the fixed stages to enhance the effect...?
The most common commercial phase shifters stop with four stages (such as the MXR Phase 90, Univibe) or six stages (MXR phase 100).

R.G. has been known to make mistakes. MXR advertises the Phase 100 as a ten stage phaser and..the schematic verifies this.
...although the effect is noticeable, it's much better if the notches move up and down in frequency.
If the Phase 100 only incorporated the six swept stages...it would not sound quite the same. The fixed stages reinforce the notches.
zaphod wrote:wouldn't it make more sense if different RC values were used in each of the fixed stages to enhance the effect...?
Actually...just the opposite. Some companies use 1% resistors or "match" the R/C components in phasers.
There are reasons why MXR did what they did. The factory documents state that the vactrols are matched. (matching the resistance slope) Vactrols are not inexpensive...you have to "weed" through quite a few to find "matching ones. Vactrols also take up quite a bit of room. Much more than a few caps and resistors.

User avatar
zaphod
Information
Posts: 7
Joined: 19 Feb 2015, 01:08
my favorite amplifier: Marshall 18W
Has thanked: 3 times

Post by zaphod »

armdnrdy1 wrote: MXR advertises the Phase 100 as a ten stage phaser and..the schematic verifies this.
Yes, I've read MXR's description and seen the schematics. But then I suddnly found myself questioning what's really going on. Just call me a cynic if you like :)
armdnrdy1 wrote:If the Phase 100 only incorporated the six swept stages...it would not sound quite the same. The fixed stages reinforce the notches.
And that's possibly where the answer lies. Maybe they want one set of fixed frequency notches with another set sweeping across them.
armdnrdy1 wrote:...just the opposite. Some companies use 1% resistors or "match" the R/C components in phasers.
Yes, I get that. However, that's a somewhat different matter from deliberately using different phase shift filter values to get a wider spread of notches across the frequency spectrum, for example, like in a Univibe, which is what I was suggesting.

Anyway, thanks for your patience in helping me think through what's happening here with the Phase 100. :thumbsup When I get some more time I may play around with switches to try and hear the difference with the fixed stages in and out of the signal chain. Maybe also add a Phase 90 switch option, with only four phase shift stages engaged.

User avatar
armdnrdy1
Breadboard Brother
Information
Posts: 104
Joined: 07 Oct 2012, 00:11
Has thanked: 10 times
Been thanked: 67 times

Post by armdnrdy1 »

As the number of phase stages increases, so do the number of notches...but...there are limitations to what sounds good.

When the control voltage is varied, it causes a variation of resistance in the R/C...which shifts the notches up and down the frequency spectrum.

Having fixed stages is like having another reference besides the dry signal. If you lift the dry signal..you have a different animal...not a phaser.

User avatar
Fender3D
Cap Cooler
Information
Posts: 573
Joined: 08 Aug 2009, 02:40
Location: Helltaly
Has thanked: 121 times
Been thanked: 153 times

Post by Fender3D »

[quote="zaphod"]
Yes, I've read MXR's description and seen the schematics. But then I suddnly found myself questioning what's really going on. Just call me a cynic if you like :)

What's going on is that you have 2 more notches with the 4 added stages...

You can't count how many notches you have only by ear... but you clearly can hear the difference, if you only bother picking up your axe and play (remember F. Zappa?)
You shouldn't worry that 2 or more stages are static, you should look how they inter-react with the moving ones

User avatar
zaphod
Information
Posts: 7
Joined: 19 Feb 2015, 01:08
my favorite amplifier: Marshall 18W
Has thanked: 3 times

Post by zaphod »

armdnrdy1 wrote:As the number of phase stages increases, so do the number of notches...
If the RC values are the same for each stage, then I think the number of notches will remain the same, but they will be deeper and more emphasized.
armdnrdy1 wrote:If you lift the dry signal..you have a different animal...not a phaser.
Very true. You need the dry signal mixed with the shifted one to get the notches occuring. Then the LFO lets you sweep those notches to get the swirling effect.
Fender3D wrote:You shouldn't worry that 2 or more stages are static, you should look how they inter-react with the moving ones
Yes, that's in fact how I'm beginning to understand it, with the help of the good people on this thread. :)
Fender3D wrote:... but you clearly can hear the difference, if you only bother picking up your axe and play (remember F. Zappa?)
This is why I think a little experimentation with switching the fixed stages out and in should be worth trying. I'm also keeping in mind that the phasers that most people rave about are true four-stage phasers, with no fixed frequency stages. So it might be neat to be able to switch between Phase 90 and Phase 100 modes. The PCB is a more recent rev though, and all the components are really packed in compared with an original script logo pedal. So it looks like it will be a bit of a PITA to work with.

Not quite sure about the Zappa reference though. I'm afraid I only remember a couple of his songs.... :oops:

User avatar
Fender3D
Cap Cooler
Information
Posts: 573
Joined: 08 Aug 2009, 02:40
Location: Helltaly
Has thanked: 121 times
Been thanked: 153 times

Post by Fender3D »

zaphod wrote: If the RC values are the same for each stage, then I think the number of notches will remain the same, but they will be deeper and more emphasized. ...
Nope, the number of the notches depends on the stages number or type (two 1st order filters make 1 notches such as one 2nd order filter does), if you check usual phaser schematics, you'll notice they use the same Rs andCs each stage..
zaphod wrote:...This is why I think a little experimentation with switching the fixed stages out and in should be worth trying...
Look for LTSpice and learn how to simulate stuff. CTRL+C and CTRL+V are really usefull when you need stages multiplying :)
zaphod wrote:...Not quite sure about the Zappa reference though. I'm afraid I only remember a couple of his songs.... :oops:
Lol
He said:" Shut up 'n' play yer guitar"

:mrgreen:

User avatar
zaphod
Information
Posts: 7
Joined: 19 Feb 2015, 01:08
my favorite amplifier: Marshall 18W
Has thanked: 3 times

Post by zaphod »

Fender3D wrote:.... the number of the notches depends on the stages number or type (two 1st order filters make 1 notches such as one 2nd order filter does)
Aha, now that makes a lot of sense from an EE perspective! :idea:
Fender3D wrote:if you check usual phaser schematics, you'll notice they use the same Rs andCs each stage...
Yes, I've pretty much studied them all, and seen that. :) However, I had always assumed the idea was just to get deeper notches. Now I need to rethink what a Univibe is doing.
Fender3D wrote:Look for LTSpice and learn how to simulate stuff. CTRL+C and CTRL+V are really usefull when you need stages multiplying :)
I already have LTSPice, but ultimately I want to hear how it sounds. :)
Fender3D wrote:Shut up 'n' play yer guitar"
I couldn't agree more! :lol:

User avatar
Fender3D
Cap Cooler
Information
Posts: 573
Joined: 08 Aug 2009, 02:40
Location: Helltaly
Has thanked: 121 times
Been thanked: 153 times

Post by Fender3D »

zaphod wrote: Yes, I've pretty much studied them all, and seen that. :) However, I had always assumed the idea was just to get deeper notches. Now I need to rethink what a Univibe is doing.
Remember, we're talking about ALLPASS FILTERS.
Raising the filter's order will raise the Group delay, then, if you cascade several filters, the higher delay will lead to a different notches position (when mixed back with dry signal).

You'll obtain deeper notches by selecting and matching caps and resistors.

By shifting the filters' fc you'll add different Group delays leading to a more complex notches position.

That's why I suggested LTSpice...

check this:
http://www.ti.com/lit/ml/sloa088/sloa088.pdf

User avatar
zaphod
Information
Posts: 7
Joined: 19 Feb 2015, 01:08
my favorite amplifier: Marshall 18W
Has thanked: 3 times

Post by zaphod »

Heh, reminds me of Circuit Theory classes from long ago. Most of the intervening years I've been working more on the dark side, with "1"s and "0"s, rather than poles and zeroes, although I've also been designing and building tube amps in my spare too. :-)

User avatar
swisher
Information
Posts: 5
Joined: 08 Jan 2017, 22:23

Post by swisher »

Can I use different value of 15uF tantalum capacitors? 10uF or 22uF?
Mesa Boogie TriAxis clone:
http://mesa-triaxis-clone.cba.pl/

User avatar
Fender3D
Cap Cooler
Information
Posts: 573
Joined: 08 Aug 2009, 02:40
Location: Helltaly
Has thanked: 121 times
Been thanked: 153 times

Post by Fender3D »

2 out of the 3 15uF caps are just DC filters, use whatever value you have (higher is better).
the third 15uF sets the LFO speed: rise it and LFO will "swoosh" slower, the opposite if you lower it.
Find your best compromise here.

BTW electros value tolerance is usually so high that you may just pick the nearest value...

User avatar
swisher
Information
Posts: 5
Joined: 08 Jan 2017, 22:23

Post by swisher »

Thanks! :wink:
Mesa Boogie TriAxis clone:
http://mesa-triaxis-clone.cba.pl/

User avatar
Seiche
Old Solderhand
Information
Posts: 2526
Joined: 01 Nov 2010, 00:16
Location: Germany
Has thanked: 231 times
Been thanked: 189 times

Post by Seiche »

I have a current MXR Phase 100 on my bench that the customer complains "phases less than it used to". What could be the problem here and how can I make sure it "phases as much as it can"?

It sounds fine to me, albeit maybe a little soft, but I've never played this effect before. How do I even begin to troubleshoot this?

User avatar
lolbou
Old Solderhand
Information
Posts: 2617
Joined: 18 Nov 2008, 21:38
Has thanked: 308 times
Been thanked: 225 times

Post by lolbou »

Did the customer tweak the trimpot?
- Are you a mod or a rocker?
- Uh, no, I'm a mocker.

User avatar
Seiche
Old Solderhand
Information
Posts: 2526
Joined: 01 Nov 2010, 00:16
Location: Germany
Has thanked: 231 times
Been thanked: 189 times

Post by Seiche »

I don't think he did because the pcb is turned upside down on the new version. One would have to unscrew the pots (which i did) and i don't think they did that. The trim pots also look pristine. The pedal is not with me anymore because i don't think there was anything wrong with it and because they wanted to sell it and not let me mess with the trim pots to get it to phase more.

User avatar
Smoky Black
Information
Posts: 15
Joined: 12 Nov 2007, 06:10
Been thanked: 9 times

Post by Smoky Black »

Image

Hello,

I’m adding this factory schematic to the thread.

Aside from the 15k/18k feedback resistor, the mixing circuit - the early scripts look like they have two 1 uf capacitors in the mixer, there also seems to be resistor changes in what looks like the lfo section? I’m talking about r2 and r6 in the schematic I just posted.

R2 seems to have values of 2.4k or 3.9k in the schematics posted earlier in the thread. R6 in the schematic I posted is 22k, but is 56k in the earlier schematics. I recently acquired an ‘81 block logo with led and that resistor is 43k. I’m wondering what those resistors do? I’m kind of curious to swap to 56k, but also wary of throwing off the center freq trim.

Thanks!

User avatar
george giblet
Resistor Ronker
Information
Posts: 404
Joined: 30 Dec 2008, 17:27
Has thanked: 206 times
Been thanked: 309 times

Post by george giblet »

R6 in the schematic I posted is 22k, but is 56k in the earlier schematics. I recently acquired an ‘81 block logo with led and that resistor is 43k. I’m wondering what those resistors do? I’m kind of curious to swap to 56k, but also wary of throwing off the center freq trim.
It's a while since I've looked at this closely but that circuit is very sensitive the part values. One cause is the small value of the 20 ohm resistors 20mV change in voltage at the base means 1mA change in LED current. If you play with R6 it's likely to mess with the centre frequency. If you pick a collector resistor then adjust the trimpot there is a certain voltage on the collector of Q1. If you change the collector resistor the trim pot would need to be readjusted to get the same voltage. As a first approximation the net result is very little change in behaviour. The reason they changed the collector resistor is likely to be due to differences in transistors and they found the trimpot would need to me maxed out in some circumstances. Tweaking the collector resistor would recentre the trimpot in production. I can't remember the fine details but that's probably the reason. I seem to remember the values on one of the versions barely worked correctly.

User avatar
Smoky Black
Information
Posts: 15
Joined: 12 Nov 2007, 06:10
Been thanked: 9 times

Post by Smoky Black »

george giblet wrote: 19 Jan 2021, 00:35
R6 in the schematic I posted is 22k, but is 56k in the earlier schematics. I recently acquired an ‘81 block logo with led and that resistor is 43k. I’m wondering what those resistors do? I’m kind of curious to swap to 56k, but also wary of throwing off the center freq trim.
It's a while since I've looked at this closely but that circuit is very sensitive the part values. One cause is the small value of the 20 ohm resistors 20mV change in voltage at the base means 1mA change in LED current. If you play with R6 it's likely to mess with the centre frequency. If you pick a collector resistor then adjust the trimpot there is a certain voltage on the collector of Q1. If you change the collector resistor the trim pot would need to be readjusted to get the same voltage. As a first approximation the net result is very little change in behaviour. The reason they changed the collector resistor is likely to be due to differences in transistors and they found the trimpot would need to me maxed out in some circumstances. Tweaking the collector resistor would recentre the trimpot in production. I can't remember the fine details but that's probably the reason. I seem to remember the values on one of the versions barely worked correctly.
Thanks, George. I appreciate your reply, it makes a lot of sense!

User avatar
DjPsychic
Breadboard Brother
Information
Posts: 147
Joined: 07 Feb 2020, 21:35
Has thanked: 85 times
Been thanked: 48 times

Post by DjPsychic »

Smoky Black wrote: 18 Jan 2021, 18:27 Image

Hello,

I’m adding this factory schematic to the thread.

Aside from the 15k/18k feedback resistor, the mixing circuit - the early scripts look like they have two 1 uf capacitors in the mixer, there also seems to be resistor changes in what looks like the lfo section? I’m talking about r2 and r6 in the schematic I just posted.

R2 seems to have values of 2.4k or 3.9k in the schematics posted earlier in the thread. R6 in the schematic I posted is 22k, but is 56k in the earlier schematics. I recently acquired an ‘81 block logo with led and that resistor is 43k. I’m wondering what those resistors do? I’m kind of curious to swap to 56k, but also wary of throwing off the center freq trim.

Thanks!
I just picked up a early block with no LED. Any idea how close to early scripts these are?

I actually haven’t received yet, but will open it up when it arrives.

User avatar
Nicale19
Information
Posts: 26
Joined: 10 Aug 2019, 09:11
Has thanked: 6 times
Been thanked: 3 times

Post by Nicale19 »

Hi guys

I write here instead of creating a new post because object title is perfect already.

Unfortunately I find myself a phase 100 with burnt ic, I had to remove them and mount sockets, but I burned some tracks

Could someone show me a layout for this model so as to check all the tracks?
Attachments
Screenshot_2022-10-31-07-20-31-68_99c04817c0de5652397fc8b56c3b3817.jpg
Screenshot_2022-10-31-07-20-23-28_99c04817c0de5652397fc8b56c3b3817.jpg

Post Reply