Zigisg Portable Amp

[Deleted]

Oh well, at least there's a reason why I thought it sounded different! I'm always more wary nowadays in case my head is telling that there will be. There's a lot of pressure on subjective listening reporting.

It is extremely cheap as well, so we're not looking at anything high end.

[Deleted]

I think I would be a little concerned as to Frans comment below;

input coupling caps are taken out.
positive: less parts, less noise (as the input resistance is smaller)
negative: DC on the input will be amplified and can silently destroy headphones if a DC is present.
This will manifest itself in a scratchy pot b.t.w. so there is a 'check' possibility for DC.
Personally I wouldn't want the caps out of this circuit as there is no protection whatsoever.

No biggy if you toast a pair of cheapie 'phones but if you have something like the BD DT1350 or Senny IEMs go into meltdown, you is gonna get a bit pissed mefinks

[Deleted]

Well, if the pot goes scratchy, the amp goes.

What are the chances of something going wrong so that DC goes out to the headphone?

[Deleted]

Slim, I hope!

Problem is, now you know it is a possibility.......

[Deleted]

So following the same set of rails then, does that mean taking the option of bypassing the caps on my Sunrise Amp which does result in a "scratchy pot", run the the risk of dc reaching and posibly destroying my HP`s then!!

Mick.

[Deleted]

It sounds like removing the caps cleans the sound a bit but then you remove protection against the dreaded DC attack!!

I guess it depends on what could cause a DC attack .......

[JohnnyBlue]

I seem to remember getting into a real panic about DC frying my headphones when I'd just built my first (and, as it currently stands, last!) DIY CMoy HPA. I asked for advice here and elsewhere and was warned about my kit's lack of output protection.

In this case it was to do with the possibility of the two rechargeable batteries discharging at different rates and thus presenting a big DC offset to the HPs, so don't know how this could apply to the single-battery Zigis.

In the end, because I was told that I'd hear the amp distorting long before it'd fry my HPs, I decided to do f*ck-all about it (a not atypical response of mine, I'm afraid) and have never had a problem, but I guess the risk is there...

EDIT:
There's an irony: I've just noticed that the warning about the dodgy circuit on my DIY Cmoy kit was given by none other than...

...Zigis!

[Deleted]

I will explain the scratchy pot thing.

If the pot goes scratchy of old age so to speak, it is no problem and no DC will occur at the output.
The 'problem' only occurs when the source that is connected to the amp has a certain amount of DC on it.
Believe me it exists, although not many gear has that problem.
In most cases every source already has a coupling cap n the output path or somewhere in the sound path.
Portable DAP's/phones etc do not have DC on it's output so little worries there.

With the 'old' type this DC could not reach the amplifier (opamp) which is set to have a certain gain.
This gain (in this case) is exactly the same for DC as for AC signals.
The cap blocks this DC (it cannot pass) but AC passes so in case a source is used that has a certain amount of DC superimposed on the output signal this DC component is take out and only the music (AC) is amplified.
This new design is all DC coupled so if for instance the source has a 0.1V DC on it and the amp has a gain of 5x the DC on the output will be 0.5V. This will cause the membrane to move to one side but the audio is also amplified so you don't hear the DC at all and won't know it's there. If that DC component is say 1V and the amp is set pretty loud and the AC component (the music signal) is 0.1V you can end up with 5V DC and 0.5V AC signal. Also in this case you will not hear this DC component (well perhaps it doesn't sound quite right)
The DC will fry the voice-coils without you knowing it.

This is true for most DC coupled amps but the more expensive ones have either DC servo or a monitoring/protection circuit that switches off the load and thus prevents damage.

Now, what has all this to do with a scratchy pot.
IF a volpot has DC on it and you rotate it, the little 'steps' it makes produces audible noise.
a scratchy noise as if the pot is faulty but sounding slightly different.
When the DC is not present and the same pot is rotated you won't hear it.
Hence my remark... a scratchy pot may be an indication there is a DC component on the source.
rare but possible.

Any C'moy can fail and apply a DC voltage on the voice coils, in fact normal amps can also fail and do this as long as they are DC coupled and don't have any protection.
Opamps usually are short circuit protected and the chance of them failing is extremely small so no need to worry about that.

Now for the Sunrise.
This amp's output is AC coupled by the big 2200uF caps so any DC will be blocked and can't reach the headphone unless this output caps fails and shorts. (very small chance of that ever happening)
In fact these amps are build on the fact it has a DC and the output caps are essential in this design and cannot be removed .

In case of the Sunrise and the DC coupled input (caps bypassed) the leakage current of the grid causes a small DC to fall over the volpot.
So when there is no music and you rotate the volpot you hear a scratchy sound.
If the exact same sound is heard when adjusting the volpot of the new Zigis you can be fairly sure there is a DC signal on it.
In case bipolar opamps are used in this amp you might hear a very faint 'scratchy noise' when adjusting the volume without any signals attached. Perfectly normal and harmless.
So for the Sunrise a scratchy pot may mean:
There is DC on the input of the amp (in this case do not bypass the input caps).
The tube is leaking grid currents (tube dependent) and the input caps are bypassed.
In both cases the headphones are perfectly safe even without any protection circuits.

As for the latest entry.
The zigis amp doesn't have the uneven battery problem because of the used rail splitter.
This is the advantage of using a rail splitter.
It also has disadvantages of increase return resistance and current limiting that is a-symmetrical.
In the O2 my evil twin has addressed this 'uneven battery voltage problem' by adding a circuit that cuts the power to the amp on both power rails when this occurs. The advantages of NOT having a railsplitter can be read in some of my evil twins essays.
Railsplitters are sometimes called virtual grounds or 3rd channel.

Most opamps do behave well even when one power supply rail is gone so Zigis's warning won't always apply but CAN be problematic with some opamps.

[Deleted]

I don't know how risky it actually is or whether it's just one of those possibilities like touching the live wire on a plug because it's come loose ...... (ie very rare)

It does seem to have an effect on the sound and Zigis did say that he had changed it in order to make it cleaner sounding, which it is.

I was starting to doubt my lugs when you pointed out the (obvious) missing caps and thought that it was the op amp I was hearing. Differences are small but detectable with the K601.

I am surprised at such a simple little circuit sounding so good. The only weird thing I detected was with nose cancelling headphones (Senns) it gives out a mains buzz until I touch the chassis of the player. (Ipod) It doesn't do that with ordinary headphones though. It's very quiet in fact.

I seem to remember the Senns did that with another amp I had so perhaps they are prone to buzzes etc.

Well, if my eyes light up with DC, I'll put it here!!

[JohnnyBlue]

Ah, very interesting Frans, thanks for the explanation.

I've noticed, on occasion, with one or two of my HPAs (can't remember which, but I'm pretty sure one is the PA2V2) that I get the 'scratchy pot' noise when the amp is first turned on: it settles down (i.e., stops being noisy) after a fairly short time, when the music starts to play.

[Deleted]

You realize the music signals that reach your ears probably have passed dozens of coupling caps of much worse quality than the ones he previously used.
Also note that he used to apply those caps for a reason as the amps circuit has always been the same and could have made it this way from day one.
He may be using it as a selling argument and saves on manufacturing costs by omitting these caps.

The humming part may indicate ground-loops or oscillations.

The circuit is simply an opamp with 2 resistors.
There is no magic fairy dust or huge know how, simply the most standard of all standard applications and this circuit is mostly even given in the datasheets of most opamps.
he does sell them at very good/reasonable prices and that's a big plus for VFM !

The same opamps circuit is used in almost any device and the OPA2134 is widely used in a lot of audio gear (higher priced ones otherwise the 5532 is mostly used) so if this would sound bad everything would sound bad.

The PA2V2 only has 3V power supply voltage and a rather bad performing circuit.
yet even that amp doesn't sound bad, it merely runs out of steam with inefficient or high-impedance headphones.


a scratchy pot for a short time on switch on indicates a relatively large input coupling cap is used (deep reaching frequency response) so when this cap 'charges' a current flows making the pot scratchy till the cap is fully charged. After that time there is no more DC current passing through the pot and thus the scratching sound fades away and stops.

[Deleted]

Thanks Frans.

The 'scratchy pot' noise I understand because I have heard it before. It's not the same as noise on the vol pot while it goes up and down, but like a 'pressure' noise, slightly smoother sounding than a dirty pot. (?) It's kind of like the pressure changing inside the earcups and not the normal 'pink/white noise you get with a dirty pot.

It sounds slightly 'smoother'? I think that's what it is.

Can anything be done to them to provide protection from DC or is it just fixed with the design?

[Deleted]

Exactly the right description.

pots that are scratchy because of dirt usually have a higher pitched noise and 'intermittend' sounds without the 'pressure' sound.

It's a design 'flaw' or should I say property.
It also has benefits and makes sure there is no (loud) switch-on pop sound or (loud) switch off pop sound heard.

a DC protection circuit is much more complicated, uses extra current to fit so it's left out.

There is a way to use coupling caps at the output (like in the G2-Sunrise e.t.c.) but electrolytic caps are said to have even worse sonic properties than the previously uses poly-something caps at the input.
This would involve extra caps that need to be able to handle at least 25V and have a value of 2200uF so will be bigger than the whole amp.
It would also give a considerable switch on/off 'pop'.

[Deleted]

Ok Frans,

Thank you, no worries then.

Mick.

[Deleted]

Reading the last few posts, I now realize I have a 'dirty' pot on my V2. I get a scratchy sound from 11 o'clock to 1 o'clock which goes away if I sweep the volco back and forward a few times.
Would it be worth dribbling a bit of WD40 (or a good audio contact enhancer - I have some Kontak somewhere) down the shaft so it trickles into the body, or is it time for a replacement?


Jeff

[Deleted]

The contact cleaners usually work miracles... for a short time... the effect wears off soon again.
How fast the problem returns totally depends on the mechanical construction/used materials of the volpot.

Cheaper pots can sometimes be disassembled, the track and wipers cleaned and put together again.
Lasts longer than a simple spray as oxidation that is wiped away comes back.
Even when oxidation preventive spray is used, the film on the surface you want treated (the wiper track) is scraped clean after some time and oxidation starts again.

For a more permanent solution replace it with a higher quality pot if possible.
In all the years I have used pots the blue velvet types have NEVER given me any trouble nor the more expensive bourns types for that matter.

[Deleted]

Thanks Frans

[Zigis]

Hello!
I reed this tread, Ian point me to, last few days. Now is time to answer some questions.
However... almost all is answered by solderdude.

Yes, input caps are removed. In theory there is some risk of DC at output, if your source have DC at output. But this is very rare, most sources have output caps or any other DC protection.
There is few very popular DIY amps/ kits DC coupled, like amb's portable M3 and others, made by hundreds DIYers. I reed Head-Fi and other forums for years and can't remember someone got real problems with it.
So, this is statistic data:) I think, there is not much to be scared about.
You know, how audiophiles say about caps in signal path - best cap is no cap:)

Anyway, tip for pot scratching sound is good. If you heard this scratching sound with source connected, disconnect input cable. If scratches disappear, there is really DC problem in your source. If scratches stay after disconnecting, it is just dirty/faulty pot.

Problem with two battery schematic and middle point ground, I talked with JohnnyBlue in other forum is different. There in worst scenario, uneven battery discharging amp itself can produce DC at output. Not with my amp.

[Zigis]

In new amp I use gain 4 .
Before I used gain 6 and come to conclusion, it is too much in almost any situations. Too small pot rotation angle is used for real music listening.
In German HIFi forum they found, even gain 2 is enough. So if anyone here have my old amp with DC socket and LM4562, amp is probably with gain 2.

As you know, new amp is with build in 5.5/2.1mm DC socket. "Fool proof" diodes are used too, this mean, if someone accidentally connect wrong polarity, nothing is damaged.

Now amp's modification with build in charging circuit are tested and soon will be listed. In this amp battery can be charged without removing, with external 12V-30V PS.
This is 16mA constant current charger, can charge 9V battery in 8-12 hours (overnight).
For charging only, you can use even cheap, noisy switch mode PS.

I think there is place for both amp's modifications.
For those, who use amp mostly on the go, rechargeable version is the best.
If someone use amp at home, with PS or external powerful battery pack, amp without charger is better, because long time PS connection can overcharge battery.
Of course, rechargeable amp user can replace battery and use amp with external PS in long run.

[Deleted]

a free tip:

To lower the internal impedance of the power supply and get a fraction more output voltage you could replace the 1N400x diode from the battery with a Schottky diode, as a rechargeable battery already has a lower voltage (8.4V) and will be lowered by 0.5 to 0.6V by a 1N400x. Shouldn't cost that much more to produce.

The current limits set by the TLE2426 (which unfortunately is asymmetric) is the downside of the railsplitter configuration and will determine the (uneven) clipping properties which also can result in an effective DC voltage on the output when clipping on low-impedance headphones.
Granted... this can be clearly heard, as distortion will be considerable, so not a DC problem in a sneaky sense anyway.

The increase in resistance of the ground (the rail splitter virtual ground) is another drawback of the circuit.
The effects are shown in my article about resistance and impedance (the effects of a higher impedance return path)
The caps in the power supply (100uF ?) can, for very short moments, supply some extra current if needed, so for very small peaks in a music signal the output power is determined by the opamp current limiter so marginally higher.

[Deleted]

Hi Zigis,

Welcome to RG. I compared the V1 with V2 and there is a difference in the clarity; especially in the top end. I was really interested in the effect of removing caps and John sent me his V1 to listen to.

I instantly heard a more extended top using the AKG K601. It surprised me how well the amp pushed the K601 since as you know, it's not an easy load and the Zigisg seems to take it in its stride.

I also realise that the chances of DC going up the chute are extremely low; especially if it's used with portable gear - so it's a theoretical chance in real life applications.

What surprises me is how good a simple circuit sounds and made me rethink other gear - perhaps we sometimes have things a little bit too complex for our own good and introduce other things as a result? A simple CMoy like this sounds really good to my ears and the power that it delivers was also a surprise.

When you hear such a simple design, you wonder if we sometimes 'over egg' the cake with some amps!!

I also think that the price level is fantastic and opens up a completely new world of properly amped sound for the Ipod kids. It's fantastic value for the sound you get and btw, I forgot to say thank you for the fast delivery too.

I've been using it exclusively since it arrived and really do like it a great deal.

[Deleted]

It surprised me how well the amp pushed the K601 since as you know, it's not an easy load and the Zigisg seems to take it in its stride

Actually the K601 is the about optimum load to drive with C'moys.
Because the impedance is very medium (120 Ohms) it doesn't need a lot of voltage so portable amps with 9V can drive it loud AND it draws less current than low Ohmic cans.

I couldn't resist... some math.

C'moys cannot provide high amounts of current because opamps are designed to drive 600 Ohm loads (aside from the 4556 for instance who can deliver substantially more).
This opamp(4556) is used in the O2 and Grado amp b.t.w. but won't bring any advantages here because of the used railsplitter type.

Especially in case a 30V supply voltage is used any headphones between 120 and 600 Ohm can be driven loud.
the maximum current an OPA2134 can deliver is about +/- 45mA (symmetrical clipping behavior).
The TLE railsplitter, however, can provide max +35mA and -70mA.
So in case both channels are driven at the same time and phase (as is usually the case with stereo) the maximum current is not determined by the OPA2134 anymore but by the railsplitter that must handle 2 return currents at the same time.
Since it is asymmetric the lowest current determines the max current.
So the 35mA is divided over 2 channels, which means each channel can supply 17.5mA. In practice it will be better a bit so lets say 25mA per channel.
Above these currents distortion will increase.

This is so called 'peak current' and to get to the 'RMS value' it must be divided by SQRT2.
So about 18mA_RMS per channel.
Since power = I_RMS²xR this means 40mW max into 120 Ohm.
The voltage that belongs with this power U=IxR so 2.16V_RMS.
To see if this voltage fits in 9V we multiply this RMS voltage by 2xSQRT2 and get 6V. We also need to include voltage drop at that current so can add 2V.
In total we need at least 8V to reach the maximum it can deliver in 120 Ohm.

How loud will an K601 go ?
the efficiency is 101dB with 1V on it.
The C'moy can supply 2.16V into 120 Ohm so this means a factor 2.16 higher.
When we convert this factor to dB's we get +6.7dB so the K601 at maximum power will deliver sligthly over 107dB, which is quite loud already.

Take into account the TLE is bypassed with reservoir caps that can handle short peaks (music consists of peaks) you might even reach 109dB with some distortion which is pretty loud. You won't be able to stand that pressure for long.

In higher Ohmic headphones the delivered power will be higher but only when supplied with a higher voltage which it can do.

In short a 9V fed C'moy with rail splitter TLE2462 gives maximum power/performance in a 120 Ohm headphone.
So the K601 is ideal for a 9V fed C'moy.

Of course opamps sound good (and thus also C'Moys).
large quantities of them are used in the entire signal path.
If they would sound bad every recording made and played back would sound bad.

By the way the biggest differences between the V1 and V2 is not the absence of the caps alone but also:
PCB (ground plane, which is important)
used opamp
gain setting

apples and pears.. but in case all of this translates in better SQ the upgade is worthwhile, but not only because of the absence of caps.

[Deleted]

Blimey Frans, you're going to town on the CMoys!!!

There is something really quite nice about the 'honesty' of the sound in that it feels as though it's not struggling at all to provide power to drive the AKG's but also the timbre or tone quality seems spot on.

In fact the K601 is really quite lively and seems to have more character than its normal flattish sound. The bass in particular is fast and very present on the Zigis.

It's actually a really nice sound and you feel a bit daft recommending people to stick a £30 CMoy on a K601 or even a K701!! &01 isn't quite as good imo on it.

At least what you're saying seems to fit what I'm hearing because I do get a bit insecure with my own thoughts about sound sometimes and I didn't expect this from a simple little CMoy.

Interesting what you're saying about current delivery since Grado push the CMoy in a posh box for their headphones and if I'm honest, my feeling was that the K601 in particular seems to get the most from this amp in comparison to quite a few others I've tried. (Grados included)

Plugged in, it seems to have a lot of authority actually.

Makes you wonder whether going back to a basic simple design is better for sound. While we introduce things to get rid of certain artifacts and improve other things, it kind of introduces something else at the same time!!!

This is a cheap little set up for a newbie when you combine a K601 with a £30 something amp!! Super sound!!

[Deleted]

The Grado can deliver considerable more power.
Into 120 Ohm it manages 250mW opposite 40mW of the Zigis (regardless of the power supply voltage)
The Grado can push the K601 to it's edge with 115dB !
Nearing the pain treshold.

The price difference, however, is VERY big yet the sonic qualities (right upto 40mW) will be very similar as the circuit is the same except for the virtual ground, which the Grado doesn't have.
Grado too found using a simple opamp is good enough.
Right until it was opened up people will probably have said very good things about it.
The Grado, similar to almost all C'moys has input coupling caps.
Those in the RA1 are very expensive ones though.

Fortunately for me, being a VFM guy first, I would buy a Zigis (if I couldn't make one myself).
The Neco portable might be better looking/have more power but is also more expensive but won't play high-impedance headphones equally loud.

Affordable and good enough for headphones above 80 Ohms it certainly is.
For driving low impedance headphones I would choose a design without the rail splitter, or one with another type of rail splitter.

Especially for 300 to 600 Ohm headphones on a 30V (or even 24V) power supply it will be hard to beat VFM and output power wise.
in 300 Ohm it can deliver about 100mW !

[Deleted]

Your knowledge never ceases to amaze me Frans, truly mind blowing, i`m sure you should be designing time machines or the like.

Brilliant.