Intersound IVP layout

Started by jimladladlooklike, October 31, 2020, 01:22:32 PM

Previous topic - Next topic

jimladladlooklike

Also, I just noticed that before and after the t former there are two separate 4558 ICs, both of which are only having one half used. Would it be possible to use both halves of one 4558?

Thanks

iainpunk

ow, yeah the effects send is the new output, and that is the output amp, since you probably dont want to stack buffers, that's a waste of components, power and board real estate.

i recommend merging halves, independent of the circuit, there are very few cases where you actually need to keep the opamps in their original pairing, most of the time you can get away with shifting that around a bit, in this case you can too.
if you can work out a layout, i recommend using quad opamps, they are generally cheaper per opamp, and it saves a bit of power management since you only have half the amount of chips.

cheers, Iain
friendly reminder: all holes are positive and have negative weight, despite not being there.

cheers

jimladladlooklike

Okay got it.

Makes sense about the quad op amps! Wish I'd thought to ask sooner as I've done most of the layout with dual op amps. Chalk it up to experience I guess.

Thanks a bunch for your help. With any luck I will have the full layout ready to post soon.

One more thing I'm curious about is whether I can avoid having to use 240v mains power and just use a 9v to 16v converter?

Cheers

PRR

OTOH quad opamps lead to a lot of wiring in a small area and increased leakage between stages. (R.G. may have notes on this.)

Neither way is wrong. Duals are very popular.
  • SUPPORTER

jimladladlooklike

Yeah I suppose that also makes sense, thanks.

I'm up to the point in the layout where I need to place the xfrmr. Could anyone tell me which 4 wires to use if I use the following xfrmr?

https://www.mouser.co.uk/ProductDetail/42TM018-RC/

Cheeeeerrs again.

jimladladlooklike

Wait never mind I think I've got it. I'm presuming that wires 2 and 5 are centre taps? So I'm also assuming that I need wires 1, 3, 4 and 6...

iainpunk

i haven't had much experience with chare pumps and changing 9v to higher voltages, but i don't think the circuit will draw a lot of current so that might not be a problem.
something like this could be something to experiment with:
https://www.ebay.co.uk/itm/12V-Dual-Output-DC-DC-Converter-Step-Up-Voltage-3-3V-5V-9V-to-12V-Boost-Module-/253029625315
a nice large capacitor from both the + and - to ground is important to filter the switching noise.

cheers, Iain
friendly reminder: all holes are positive and have negative weight, despite not being there.

cheers

ElectricDruid

Quote from: jimladladlooklike on February 15, 2021, 04:17:47 PM
Okay got it.

Makes sense about the quad op amps! Wish I'd thought to ask sooner as I've done most of the layout with dual op amps. Chalk it up to experience I guess.

Opinions differ, but I prefer duals for layout because then all the op-amps "point in the same direction". This helps signal flow around the board. On a quad package everything points out from the middle, which I find more difficult to lay out. Plus the power pins are reversed with respect to virtually everything else, which makes it awkward to put them on a board with other parts (unless you add them in upside down, which I also don't like).
If you play around with both, you'll find which you prefer. Depends a bit on your design style, I guess.

Quote
One more thing I'm curious about is whether I can avoid having to use 240v mains power and just use a 9v to 16v converter?

How much current does it draw? Usually voltage boosters are ok as long as the current drawn if fairly light. Big loads make that a lot harder. But doesn't this thing have a +/-16V bipolar supply? In which case, you'd be doing a lot of gymnastics to get that from a +9V input...not impossible, just tricky.

You could think about running from an 18V input and then just adding a rail-splitter op-amp to give you a +/-9V supply. That'd let you use the circuit without making any serious modifications, and 9V isn't so massively different from 16V that you'd notice much. At least, not in this circuit for this application. Yes, you lose a bit of headroom (18V vs 32V), but if we're turning it into a pedal, we don't want mixing desk levels anyway.



jimladladlooklike

Quote from: ElectricDruid on February 16, 2021, 02:21:23 PM

How much current does it draw? Usually voltage boosters are ok as long as the current drawn if fairly light. Big loads make that a lot harder. But doesn't this thing have a +/-16V bipolar supply? In which case, you'd be doing a lot of gymnastics to get that from a +9V input...not impossible, just tricky.

You could think about running from an 18V input and then just adding a rail-splitter op-amp to give you a +/-9V supply. That'd let you use the circuit without making any serious modifications, and 9V isn't so massively different from 16V that you'd notice much. At least, not in this circuit for this application. Yes, you lose a bit of headroom (18V vs 32V), but if we're turning it into a pedal, we don't want mixing desk levels anyway.



Oh right, thanks for pointing that out. It hadn't dawned on me that I would need to get both +16v and -16v from +9v... Will do some research on how.

I don't know how much current it draws, honestly. How would I calculate that? Watts/Volts? How do I find out the wattage? Excuse my ignorance.

Thanks

iainpunk

an educated guess: it uses less than 150mA

you might also want to look at the transformer voltage/freq rating. i expect the transformer to be slightly overdriven, but at a lower voltage, you might not get that, so trying lower voltage transformers might be a good idea here.

cheers
friendly reminder: all holes are positive and have negative weight, despite not being there.

cheers

jimladladlooklike

Right, so I think it may be best to abandon the idea of powering with a +9v wall wart.

Instead, would I be able to power it with +32v DC and split the voltage somehow?

iainpunk

you can probably get away with only using 18v to 24v, i think 32v might be overkill here, the gain isn't that high, and the circuit will most probably work just fine on + and - 9v or 12v.

i suggest looking in to charge pumps and other boost converters like the Ebay link i posted earlier, this makes 9v in to a higher voltage

cheers, Iain
friendly reminder: all holes are positive and have negative weight, despite not being there.

cheers

jimladladlooklike

#32
Okay, thanks again. Still getting to grips with power supplies and what different circuits require/whether they are affected by a differing voltage.

But, I have finished the vast majority of the layout... I think.

A quick note: I like to drill a larger hole through one of the holes in the strip board to create a trace cut which I then tidy up with a tool afterwards. The DIYLC software only allows lines to be drawn between said holes, so each of these lines represents a hole that will be drilled to the RIGHT of it.

Here we have the input preamps. There are two input jacks on the IVP, both with a lo/hi impedance switch on them. I've simplified it slightly so there is one low and one high, neither of which you can change with a switch. (Also, with regards to the PS decoupling network, I'm not sure I will need this if using a charge pump? I'm guessing this is to eliminate DC ripple? I left it in just in case it's necessary.)



Following that we have the elaborate EQ section:



And finally the tube voice part:



As you can see, especially with the EQ, it's a lot of strip board -but I want to put this in some kind of big chassis and use it as an "always on" kind of thing as part of my rig (which it's looking like I will never use again at this rate, but whatever).

Any advice whatsoever would be truly appreciated, likewise if anyone has the time to check it over as per the schematic. I've checked it myself but it's late and my eyes hurt!

PS layout to follow, once I manage to get my head around what to do. Every day is a school day!

EDIT: The resistor on from the emitter of the 2N5210 to -16V on the left of the TV circuit layout is meant to be 36K, oops.

ElectricDruid

Quote from: jimladladlooklike on February 19, 2021, 06:29:11 PM
A quick note: I like to drill a larger hole through one of the holes in the strip board to create a trace cut which I then tidy up with a tool afterwards. The DIYLC software only allows lines to be drawn between said holes, so each of these lines represents a hole that will be drilled to the RIGHT of it.

DIYLC has a "trace cut" option on the "Boards" menu which is the best way to represent a trace cut on veroboard. It looks exactly like drilling a track out with a wide drill bit. Lines don't work well because they're not the way it was designed to be done.

jimladladlooklike

Seems I'm using a very old version... will download the new one now!

jimladladlooklike

2nd draft:

2n4401 + 2n4403 on EQ board swapped positions and diodes changed accordingly.

Pin 2 of the furthest right 4558 on EQ board now connected to lug 3 of all freq level pots.

Traces now cut using proper tool.

xfrmr added for more accurate size.







jimladladlooklike

Hey again,

I've been looking into this power supply situation. I still want as close to the +/-16v bipolar supply as possible, but admittedly don't have the knowledge to get that from and +9v unipolar wall wart. I have come across this schematic for an op amp voltage splitter which utilises an LM741.



Theoretically, if I found a 32V AC adapter (or an adjustable one that had that setting or, say, +30V) could I use this circuit. The max supply voltage for the LM741 is +/-22V but does that mean it could essentially handle max +44V from a unipolar supply?
Another thing I'm concerned about is the germanium PNP transistor (MJ2555). Could I replace this with a silicon tranny?

duck_arse

that transistor should, most likely, be marked MJ2955, which is the compliment to the 2N3055, both are silicon. the same transistor pair comes in different packages with different prefixes, eg TIP3055 or MJE3055, probably others I can't remember. that transistor pair would have been selected for their MASSIVE-ish current handling, probably way more than you would need here.
" I will say no more "

Rob Strand

QuoteTheoretically, if I found a 32V AC adapter (or an adjustable one that had that setting or, say, +30V) could I use this circuit. The max supply voltage for the LM741 is +/-22V but does that mean it could essentially handle max +44V from a unipolar supply?
Another thing I'm concerned about is the germanium PNP transistor (MJ2555). Could I replace this with a silicon tranny?
The transistor is actually an MJ2955 which is the PNP compliment to the 2N3055.

You don't need such large transistors for that circuit.
[Edit: duck_arse just posted the same]

Perhaps you could consider an AC input with a voltage doubler feeding a voltage regulator,

http://www.electricalbasicprojects.com/voltage-multiplier-circuits-half-wave-voltage-doubler/

For the positive supply you use one doubler,


For the negative supply you use another doubler except with the diodes and caps flipped.

Connect the input caps from each doubler together then to the transformer.
Connect the "through rail" on the each double together then to the other side of the transformer.

If you start with a higher AC voltage you can use this type of doubler,
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

jimladladlooklike

Cool, thanks for the suggestion, but I don't know where I'd get an AC voltage from. Do such wall warts exist? Also I'm not sure which transformer you mean? I'm trying to steer clear of the need for a PT if that's what you were suggesting. Could I use a version of the schematic I posted but with more apt transistors?

Cheers