A request by KG who also provided lots of useful information and a schematics for this. A bit more simple version of the DIYRE Tape Saturator circuit using op amps and there is no need for a rack circuitry.
The circuit is supposed to run on 16v / -16v so there is an onboard bipolar voltage doubler although it only gives out around 15v /-15v but it is close enough.
Info about the circuit HERE.
'
'
Awesome, thank you!
ReplyDeleteI see only 25 cuts on the layout?
ReplyDeleteNow it is 26 cuts. Forgot to add a cut between pin 1 and 8 at IC3. The cut can be omitted you want to use MAX1044 or ICL7660S.
DeleteIt's coming along nicely so far...
ReplyDeletehttps://i.imgur.com/tFo5mdq.jpg
Urgh. I have only 2x 100µF 50V caps, and at least 10 of them rated for 35V... Couldn't the 35V handle the +/-15V swing?
ReplyDeleteYes, maybe. But tolarences can be a bit off when it comes electrolyctic caps. With that in mind, 35v is a bit close to 30v...
DeleteYou can safely use one of the 35v caps for the 100uF above IC3 it only 9v before the charge pump.
If you have a spare 47uF 50v cap you could try it temporary for the spot just north west of IC1.
Thank you for your help.
DeleteI have 47µF/50V, as well as 150µF/50v : shouldn't I try with the bigger value?
I think both values will work since that cap is for power filtering.
DeleteI think you can use 100uF 35V electros for all of those except the cap off pin 5 of the LT1054. They are all seeing only the positive or negative rail of the power supply, so they only need to be able to handle 15V (25V+ to be safe) rather than the full 30V (50V to be safe).
DeleteI don't know about pin 5 though, you'd have to check that with a multimeter to be sure.
Hello,
ReplyDeleteSo, I built it, but yet it doesn't really behave the way it should : when moving either the level or the gain pot, all I hear is transparent boosting... No saturation, no compression no filtering whatsoever :(
The only difference with the layout is that I used an external pot instead of the trimmer, with lines S & U wired to lug 1 & 3 (I didn't wired lug 2 to line T since there is a bridge on the veroboard already).
I will investigate further.
Hi KG, did you connect Lug 2&3 together? Or lug 2 to Line-U?
DeleteIt sounds like you're not getting a big enough voltage swing to clip the diodes - which need probably >1V to really start doing their thing. If you're plugging a guitar in you're probably not going to get there without increasing the gain of IC1, as this circuit's set up out for line level (0.5V), not guitar level (0.1V).
DeleteYou can try increasing the gain before the diode ladder by replacing R3 with a higher value - try 50k or 100k. Better yet, test it with a pot, then solder that resistor value in once you've found what works.
Thank you for the tip.
DeleteI am not using a guitar : i made the test using a drum machine, and also a buss from my mixer.
I will try with a higher value for R3.
As RENRUT said, make sure you've connected lugs 2&3 together as well, as otherwise that pot will just be a 50k resistor. You're most likely to hear some clipping with that pot turned down, so test it at ~10-25%.
DeleteYou may also want to check the voltages your ICs are getting just to be sure.
Hello, further reporting :
ReplyDelete- I made the connection between lugs 2&3 (stupid me!), and now I can hear the compression/saturation effect. It's indeed much, much more obvious at the very beginning of the pot rotation, so I guess this is the reason why I couldn't notice it before.
- I also increased the size of R3, which helped hearing the effect processing.
So, it works, and it's a verified layout, but imho this circuit might deserve further tweaking to behave better in a wider range of situations :
- First, I don't really get the relationship between the INPUT and the GAIN knobs. Maybe i'm missing something (or their labelling is unusual), but this thing is insanely LOUD, and both GAIN and INPUT make it louder.
If I look at the schematic, it seems that the GAIN is post-diode processing, so to increase input gain, we should use INPUT instead?
So, I think I will be adding a passive output VOL pot to tame things down, but i'm wondering if we really need two pots, and if the "real" gain pot is not the "input" pot?
- Also, while pushing the processing into the red creates some nice saturation, the process creates a lot of high harmonics. That's expected, of course, but it seems that the de-emphasis stage doesn't really compensates for the pre-emphasis.
Is it because i upped R3 to 100K, and should change R8 accordingly?
And would swapping R8 or R9 for a pot allow to make a low pass TONE?
Anyway, if i may need to tweak the circuit a bit to make it suit my gear and workflow better, i'm very glad and grateful for this layout (which I was eyeballing for a while) !
Nice job, I'm glad it worked out for you. Looking forward to hear how your future tweaks turns out. A master volume control sounds like a good idea for a start. But I'm afraid I can't help you out with the technical stuff when it comes to this one, the circuit is too ubfamiliar to me but I'm sure the other guys (and girls) can give assistance...
DeleteJust need to clarify some things for some updated for the layout.
Did you joined lugs 2 and 3 for both controls?
You increased one resistor to 100k. Which one?
Cheers and thanks for verifying the layout!
It looks like upping R3 to 100K would have drastically increased the high frequencies on the output. You can either decrease C2 to 0.75nF (750pF), or put a 1nF capacitor in SERIES with C2 to turn it into a 0.75nF equivalent capacitor.
DeleteThat should bring the frequency response back in line with the original.
Hi, CrackingTheShat.
DeleteThank you for your reply, and for the tip.
Yes, this is what I thought : R3 is here both for the gain and the pre-emphasis circuit, right? If we change its value, we lose the balance with the de-emphasis occuring at a later stage, right?
I will try your suggestion.
All in all, it's a nice circuit. It requires some tweaking to the gain staging, and it's cleaner than what i expected, but it's definitively in the ballpark of a "tapey" sound, and it's not a one trick pony.
Yes, changing R3 will change the balance between pre & de-emphasis, but just matching that value in the de-emphasis stage won't fix that imbalance. I simulated the circuit in LTSpice to check that changing C2 to 0.75n would fix it: https://imgur.com/a/PG1fsf7
DeleteGreen is the Pre-Emphasis frequency response & blue is the De-Emphasis frequency response, shown from 10Hz-20kHz.
[strange : I had typed a lengthy reply, but it seems it didn't get published? I'm making a shorter reply, this time]
ReplyDelete- I joined lug 2 & 3 for the external pot I used instead of the trimmer.
- I increased R3 to 100k (instead of 22k).
- Yes, a master volume would be nice, but I think I should find out how to manage the gain stages better. At the moment, the effect is very, very loud, and I think it isn't intended to be used that way. I would like to be able to drive the circuit at the input, but keep reasonable levels (because right now my output level would be permanently set to "1", ah ah!).
- There is one thing which I don't really understand in the circuit as it is : usually, the GAIN is the input drive, driving the distorsion/clipping. Here, it seems to be the pot called "INPUT" which drives the diodes. And the GAIN pot is placed after the diodes.
[Edit : while i'm typing this, I realize that we called this GAIN, but in reality, it is the MAKE UP gain, which is different : it is here to compensate for the gain loss occurring during the dynamic processing. So, in the studio world, the knobs would be called "input" and "makeup", but in pedal world, it would be called "gain" and "volume". GAIN is usually pre-processing, driving the circuit, and OUTPUT is post-processing.
So, you adjust input gain using the "input" (we should call it GAIN), and you dial your output level using the "gain" (which we may call "volume" or "output").
The knobs will be saturation (probably input) and trim, that's what usually on the DIYre 500 modules.
ReplyDeleteI'm considering building a stereo pair of these... I will read up and price up...
Couple of Qs on the build...
ReplyDelete-what does the 10K or 50K for R4 change?
-Gain 3 doesn't look like it's going to earth on the vero unless I've missed something
-1N34A are currently stupid pennies in UK and import only (builds not worth it at the moment with those). Would a 1N34As (the orange glass uns) do or is there another suitable alternative?
Oh and what's the polarity of that final 10Uf? Guessing positive facing upwards but schematic doesn't show and polarities
ReplyDeleteI'm pretty surethe 10uF are a bipolar/non polarized electroluctic cap. They are pretty common in circuits running on bipolar power.
DeleteThanks. Amy idea of a suitable ge diode alternative to 1N34A? They are currently out of stock in the UK with only inflated prices about (talking a few £ a diode at the moment!). Seen Russian D9Ks or D9Es though for a much more reasonable price, would they do?
DeleteOk thanks. Any idea on the diodes? It wouldn't current cost about £40 from within Uk or £20 plus import and long wait for 1N34 atm!
ReplyDeleteLooked at these
https://www.ebay.co.uk/itm/232244577346
You really need authentic germanium point contact diodes for this as it relies on the soft knee characteristics. Those look like schottky's, which have a similar Vf, but much harder knee.
DeletePedalhacker has a good selection of Ge diodes, but they won't be as cheap as Si diodes.
Gonna have to leave the build then... Balls.
ReplyDeleteI'm UK based so pedal hackers no good. 1N34A have suddenly became tough to buy here, I bought a bunch for a few quid last month (and used them), now they're suddenly a few quid each.
Would a Russian diode (ie D9E) be ok instead maybe?
I've had the same problem with Ge diodes, they're getting harder & harder to find (and targets for fakes). D9E diodes are point-contact Ge's, so they should work fine. This DIYS thread goes into more detail about Russian diodes: https://www.diystompboxes.com/smfforum/index.php?topic=118007.0
DeleteIn real point-contact Ge diodes, usually you can see the kinda curvy wire inside, which contacts a pellet of germanium on one side: https://www.google.com/url?sa=i&url=https%3A%2F%2Fwww.elprocus.com%2Fcrystal-diode-circuit-working-with-applications%2F&psig=AOvVaw2y3TC28gFbZDZ67LI4CVlC&ust=1620064097650000&source=images&cd=vfe&ved=0CAIQjRxqFwoTCJjN657Hq_ACFQAAAAAdAAAAABAP
If you can't see this curvy wire, or the glass is lined with orange-colored material like in Schottky diodes, you MAY not be getting a real Ge diode. There are tons of exceptions though, like black glass & milspec diodes, so your best bet is to buy from a reputable vendor.
Ok the obvious question is: do i built the opamp or the diode version??Thanks!!!
ReplyDeleteHi there,
ReplyDeleteI'm currently in the process of building this, in addition to the 8 Germanic diodes I will add 4 more diodes via switch to get a softer saturation, like on this page: http://zadvitsky.com/index.php/studio -devices / tape-saturation-compressor.html
I have an important question, why are there so many different inputs? does everyone have a different task?
This comment has been removed by the author.
ReplyDeleteThis comment has been removed by the author.
ReplyDeleteI built a stereo version of this last night (two boards into dual pots).. The 50k trimmers an external pot as well as adding two switches to add additional diode pairs. All seems to work ok so far.
ReplyDeleteUsed D9Bs as 1N34As are rare and pricey in the UK now.