Pulsar Tremolo LED & Space mods

vanessa · December 18, 2006, 02:12:38 PM

Hello all and good day! I finished my EH Pulsar build a few days ago with great success. No ticking at all and it sounds fantastic.

I wanted to share a couple of things (real quick) that helped make this a really great tremolo. First the tick that most people come across with this unit have built it from the Tonepad layout is that there is a trace error in the layout. You can see it if you look at Tonepad's schemo @ Q3's collector to R14. On the PCB Q3's collector connects to the wiper of R15. I have the schematic for the original version of the Pulsar and it's very similar to Tonepad's schemo so the fault lies in the PCB. Just cut the trace from Q3's C right before C8 and jumper that trace to R14.

There is also a treble (volume) drop due to the circuit being inherently muddy. This was easily solved by lowering the input cap from .1uf to .047 (you can lower it to taste). The only muddy problem I had after that was at it's highest speed the "OFF's" spacing seems a little wider than the "ON's". So at a high rate there (appears to be) are more OFF's than ON's so that it is hard to hear note/chord changes. You mostly hear the effect when the DEPTH control is set max.

vanessa · December 18, 2006, 02:14:04 PM

So my questions are...

Can the OFF spacing be manipulated? (How?)

Can you get pulsating LED in sync with the LFO? (How?)

gez · December 18, 2006, 02:41:40 PM

The LFO looks like an Astable Multivibrator, so it outputs a square wave. This is shaped into something resembling a triangle by the 200k resistor (coming off drain of first LFO trannie) and the .22u cap when the tremolo setting is selected, and (I should imagine) something resembling a trapezoid when the smaller cap is selected. Either that or it'll be the same waveform as the other setting, only larger in amplitude.

If you look at curves for the charge/discharge of a cap through a resistor, you'll get an idea of the waveform that controls the base of the trannie used as a variable resistor. Not very symetrical and amplitude decreases with increasing frequency, hence the problem with on/off times being slightly out.

The treble loss you mentioned was probably the result of the low input impedance of the input stage. Buffer would be a good idea for this circuit and would be a better solution that cutting the low frequency content.

As for the flashing LED. If you just want on/off blink, buffer the output of the LFO and use it to drive your LED. A simple n-channel MOSFET with gate connected to the drain of the first transistor in the LFO should work. Source to ground and stop resistor and LED from drain to V+.

PS I've never built this circuit, the above is just my initial impression.

TELEFUNKON · December 18, 2006, 03:09:08 PM

wow: all tantalums

can that be legal?

vanessa · December 18, 2006, 03:27:11 PM

QuoteIf you look at curves for the charge/discharge of a cap through a resistor, you'll get an idea of the waveform that controls the base of the trannie used as a variable resistor. Not very symetrical and amplitude decreases with increasing frequency, hence the problem with on/off times being slightly out.

So do you think if I messed with the value of the 200k resistor right after Q1 I might be able to pull that in?

gez · December 18, 2006, 03:40:58 PM

Quote from: vanessa on December 18, 2006, 03:27:11 PM
QuoteIf you look at curves for the charge/discharge of a cap through a resistor, you'll get an idea of the waveform that controls the base of the trannie used as a variable resistor. Not very symetrical and amplitude decreases with increasing frequency, hence the problem with on/off times being slightly out.

So do you think if I messed with the value of the 200k resistor right after Q1 I might be able to pull that in?

The problem is that amplitude of the waveform will always decrease with frequency. Maybe if you reduce the 200k it will help, but at the expense of depth/chop (effect will sound more prominent at slower speed settings). Plus, you may find 'on' time more prominent if you make it too small. Again, this is just speculation on my part, I'm not familiar with the circuit.

A true triangle LFO would be the solution - amplitude stable at all frequencies.

gez · December 18, 2006, 03:44:04 PM

Quote from: TELEFUNKON on December 18, 2006, 03:09:08 PM
wow: all tantalums

can that be legal?

Only between consenting capacitors...

gez · December 18, 2006, 03:46:54 PM

Quote from: gez on December 18, 2006, 02:41:40 PM
A simple n-channel MOSFET with gate connected to the drain of the first transistor in the LFO should work.

That should have read 'connected to the collector of the first transistor...'

audioguy · December 18, 2006, 04:05:23 PM

Quote from: vanessa on December 18, 2006, 02:12:38 PM
Just cut the trace from Q3's C right before C8 and jumper that trace to R14.

Which side of R14 do we jump to?
Thanks!

vanessa · December 18, 2006, 04:21:11 PM

Gez, could you point me to a true triangle LFO? Maybe one that uses a single opamp?

gez · December 18, 2006, 04:52:42 PM

Quote from: vanessa on December 18, 2006, 04:21:11 PM
Gez, could you point me to a true triangle LFO? Maybe one that uses a single opamp?

All the basic ones use a dual op-amp. Can't do a search now as I'm half watching a really interesting documentary about adoption and it's getting to a good bit. Will post later (or could someone else do the honors?).

gez · December 18, 2006, 05:02:21 PM

Just occurred to me, if you replace the existing LFO with a triangle LFO this assumes that the 200k and the caps would be dispensed with, so you're going to loose the 'pulse' setting, you'd only have the triangle. All swings and roundabouts I'm afraid. It would require a substantial redesign to get it to do what you're asking.

vanessa · December 18, 2006, 06:55:40 PM

I guess I'll mess with the 200k resistor and see if that helps a bit. I'll post my findings soon.

rockgardenlove · December 18, 2006, 07:04:56 PM

Btw, how is this pedal? I can't decide between the Small Stone, Small Clone, and Pulsar for an Xmas gift.

Meanderthal · December 18, 2006, 08:23:20 PM

Heh- then get all 3! Never heard a Pulsar, but you're gonna love the Stone and Clone!

rockgardenlove · December 18, 2006, 08:48:08 PM

I know, the Stone and Clone are kickass indeed!

I can't afford all 3.

vanessa · December 18, 2006, 09:18:13 PM

The Pulsar (original) has a bad rap around here do to a lot of people having built the Tonepad version without having done the fix mentioned above. Without doing this fix the circuit has a crazy ticking noise when in the "chopper" mode. With this fix it is a very cool tremolo effect and the chopper switch is da bomb! Even at the fastest rates (I just turn the depth down a bit) it is very, very cool (think Pink Floyd "One of These Days" off Meddle). I will be messing with some values to see if I can get the fastest speed to be "all that and a bag of chips" (not that it is not that already).
BTW: It has a very low part count and can be made with spare parts lying around. I used 2 resistors in series for some of the values (30k = 2x15k, 200k=2x100k etc.).

vanessa · December 18, 2006, 09:35:21 PM

Here is the corrected version of the PCB. I'll take it down in a bit as it is not my artwork. I just want to show the correction for reference.

rockgardenlove · December 18, 2006, 11:13:57 PM

Do the reissues by EHX have the ticking problem?
Do you know?

I'd rather play it first, but I can't find a store around here that has any.

vanessa · December 19, 2006, 12:21:32 AM

The "re-issue" is not a re-issue at all. It's a totally different circuit. Just the name is the same.

News:

Pulsar Tremolo LED & Space mods