Indeed G2 tweaks and bias adjustment

[Deleted]

Mike
I presume that you have those 100 ohm resistors connected so that they are removed from circuit when the headphones are plugged in ?
Alex

[PinkFloyd]

Mar 21, 2010 11:18:15 GMT @sandyk said:

Mike
I presume that you have those 100 ohm resistors connected so that they are removed from circuit when the headphones are plugged in ?
Alex

Yes, of course Alex!

[robertkd]

Yep Mike's wired them on the break side, Still think for the purpose they could be 1Kr or so just to make sure the output cap electrode voltage is at ground for DC.

Mike the other issue without the secondary load reflected back into the primary of the output transformer the primary winding will saturate causing the plate to work into essentially a short circuit.

[robertkd]

Mar 21, 2010 11:22:54 GMT PinkFloyd said:

Sod it.... I just snipped them off! Just realised they were NOT out of circuit .... I've made a big mistake with this one, sorry I'm off to take a cold shower and wake myself up!!!!

Krap Mike's going to freeze to death taking a cold shower

[clausdk]

My analogi with the motor was not that far off..

Thaks for taking the time to try and explain something, that advanced..

I can not say I get all of it but I do get the big picture, I hope
It makes the amp more stable and protects both the speakers and the components, so that they won't get damaged buy a sudden raise in current or surge as electronicgurus call it



I'll better stick to my leather
That is much more simple

[PinkFloyd]

Mar 21, 2010 11:30:44 GMT robertkd said:

Mar 21, 2010 11:22:54 GMT PinkFloyd said:

Sod it.... I just snipped them off! Just realised they were NOT out of circuit .... I've made a big mistake with this one, sorry I'm off to take a cold shower and wake myself up!!!!

Krap Mike's going to freeze to death taking a cold shower

Actually.... they ARE out of circuit when you plug the headphones in!.... I was looking at the pre-out and thinking it was the input..... nope I WAS correct..... as soon as you plug the phones in they go out of circuit.

I will give the 1K a try Robert, thanks.

[insomniac]

Mar 21, 2010 10:33:29 GMT robertkd said:

Nice one Frans

but I'm thinking unless they haven't included a bleed resistor that keeps the output capacitor at DC ground so you don't get pops in your headphone as you plug it in.

As for dummy loads isn't that what's between the headphone transducers

Sometime dummy loads don't give the true readings though. In this case you may wish to try instead ....






A "load of dummies"

[robertkd]

Mar 21, 2010 14:19:38 GMT insomniac said:

Mar 21, 2010 10:33:29 GMT robertkd said:

Nice one Frans

but I'm thinking unless they haven't included a bleed resistor that keeps the output capacitor at DC ground so you don't get pops in your headphone as you plug it in.

As for dummy loads isn't that what's between the headphone transducers

Sometime dummy loads don't give the true readings though. In this case you may wish to try instead ....






A "load of dummies"

Dave I couldn't let that one pass



Robert

[leo]

Mar 20, 2010 21:43:40 GMT PinkFloyd said:

By the way..... the sound quality since I have matched both channels @ 13V and fitted 2200uF coupling caps has taken a decided LEAP up in performance.... I am hearing details in Yello's "Junior B" that I have never heard before.... a definite improvement for the better.... not as "sweet" sounding, a bit more grain and grit to the proceedings.

Leo, I would like to send you a few of these Samwha VA caps, and a couple of the 3.3uF WIMA caps so you can replicate this.... I don't doubt my own ears but it is good to have a second opinion..... I don't know "why" but these Samwha caps have injected some real magic dust into the G2.

Don't worry about sending the sample G2 to Attilio, I'd like you to spend time with it, without worrying about sending it on to the next on the list..... I will send Attilio mine..... I really want you to to replicate the above and tell me what you HONESTLY think

Mike.

Hi Mike,

Ok no probs I'll try the caps before the IRF540's
At least with the Samwha's higher 130C temp rating you'll get a longer lifespan compared to the stock jobs, if they prove to be sonically better too that'll be a bonus.

I was surprised that pillar does transfer some heat to the rear heatsinks, it does get pretty warm, I'll have to measure it just to see the difference

[Deleted]

As I already stated the question if a dummy load is really needed depends on the topology (circuit design) and PCB layout.

So yes... tube amps with output transformers may need dummy loads (in fact these already do have these already fitted as seen in the boards) and when driven without a load these amps may become defective.

Fortunately I don't use transformers for audio (well Actually I did but only in my electrostatic speakers)

When you use parts that are prone to oscilation or the PCB design is crappy or the amp has a bandwith which could amplify medium wave signals or video without blinking an eye it would be wise to stop them from oscilating using a dummy load.

I personally would not want to use a design that needs a dummy load, there are other options for making an amp not oscilate without a load (boucherot filter for instance).

Mike is absolutely right in stating that it won't hurt as long as they are mounted as Mike proposes (they need to be uncoupled when a HP is connected)

the 1k for the G1/G2 is pointless as it is already there but as also said for another reason (bleeding the output cap to prevent LOUD clicks when inserting a headphone when ithe amp is already powered on.
making it 500 Ohms tottal load instead of 1000 Ohms seems pointless.
IF you are about to do it use Mike's 100 Ohms... but again not needed for the G1/G2.

I do share Robert's concern for Mike's 'antenna's'.
these antenna's are placed in the part of the circuit which is THE MOST high-ohmic AND the most susceptible for picking up and detecting RF signals.

Mike lives in a RF friendly neighborhood but a lot of users are likely not.

You know that SMPS designers DELIBERATELY put start-up electrolytic capacitors NEAR heatsinks to 'regulate' the lifespan of these power supplies ?

Mike is very right to replace the caps close to the heatsinks with 135oC caps.

I wonder why Mike discovered the amps sounded best at 16.5 V anode voltage (G1) and after Peter said 12V is sweeter these amps now need 13 Volts and not 12V ?
It is possible that the amps work better when the LM317 are NOT operated in their minimum required setting (they will be operating within parameters from 15 Volts anode voltage and up)
So people ... put your finest ears on and report your findings based on 'blind' setting of the anode voltages and when you have found 'the sweet spot' the measure the anode voltage.
If 12 - 13 Volts is the concensus then so be it.

For those that are interested in mods I will be working on Ian's device for the coming weeks and will publish when ready AND checked by Ian if these are sonic improvements.
Will do research on proper settings of andode voltage based on symetrical output signal (max headroom).
Don't expect to see anything before a month or so.

It won't consist of swapping parts but more likely radically changing the schematics/design but leaving the 'design properties' intact meaning tube and output stage (the signal path) will be the same.
It will be autobiassing but find Mike's remarks about finding the sweet spot intriquing so I will make it so you can still do this and measure the setting voltage in a safe way and for both channels at the same time.
Tube rolling will be easy... just put them in and they will always bias on the sweet spot without the need of measuring/adjusting.

As Ian already knew before it was even posted here the PS cap will be changed to a 35 Volt type as I really can't understand why all these amps have 25 Volt types (actually I can and it's called producing cheapest way possible, 35 Volt types are simply more expensive)

damn why do I keep writing these long articles..

and yes I know... Mike is good at what he does and comes up with great ideas AND good tutorials too.
Not a bad word about the guy (not sucking up !!, I couldn't care less)

No need to ask me about my mods before I published them and Ian had a good listen evaluation.
For all I know I could be making a bad sounding amp of it ...

[Deleted]

These "tails" make it a LOT easier to adjust the bias.... two multimeters is also the best way to go (with crocodile clip probes) you can set both channels up with ease and don't have to poke about.... just clamp the tails and clamp the other test lead onto the outside of the phono sockets..... all you have to do now is turn the trimpots until both multimeters show 13V...... re-adjust after one hour.

Made up extension leads as per Mikes instruction to day and have since checked The bias on mine. 7.1v left 9.2v. right with standard valve, way below optimum of 13v, it sounds a lot nicer now to my ears. Would have thought Indeed would have taken more care in at least setting the first 10 properly(no one so far has found there`s to be correct). Its a good job i don`t send my work out like that, if i did, i wouldn`t get any more to do!!.

[PinkFloyd]

Frans..... the "tails" are not intended to be "permanent" fixtures..... maybe keep them on for a couple of weeks until you decide what valve you are going to finally settle on and then desolder them. They certainly do make checking the bias a LOT easier than probing the pads.... just clip the crocodiles on and keep them clipped on until the bias procedure is complete.... a lot more convenient than poking about on the board with a probe.

I am NOT playing "follow the leader" with regard to Peter's "12V" advice..... the advice with the G1 was "16.5V" and, to be totally honest, I didn't even think of trying it @ 12 - 13V..... I will try the G1 at 13V and see how it sounds. One thing is certain to my ears though, the G2 does sound better @ 13V (to my ears anyway) maybe Leo can comment on what he prefers?

I have always used 105C capacitors but, then again, I'm not a manufacturer.... today's manufacturers don't WANT their products to last so they build in planned obsolescence to ensure you either have to replace it after a period of time or have it repaired by them.... The 85C caps in the MF amps are prime examples of this.... guaranteed to fail prematurely in that greenhouse of an amp.

The Samwha VA series caps are excellent, originally designed to be used in the automobile industry "withstanding the high temperature stresses that are associated with automotive engine compartments allow the positioning of electronic control units close to where they are applied, rather than at some distance away in the passenger compartment"

"This helps promote weight reduction and improve fuel efficiency - a constant demand for automotive manufacturers. The components are available in values from 1 to 2200°µF and 10 to 63V. They have a rated load-life of 4,000 hours at 130°C for those of diameter 10mm or above. Dimensions vary between 8 and 16mm in diameter, and heights from 11.5 up to 25mm."

These puppies will last forever! They are also good for audio use too and have quite a lossless "neutral" signature but you probably don't believe that different types of capacitors have different sonic signatures? I have you down as the type of person who would believe 1000uF is 1000uF "end of story"..... I may be wrong?

Anyways Frans, your mods will be most interesting..... The tweaks I have just published are simply "prolonging the life" of your amp and setting it up correctly type efforts.... it's a bonus that the 2200uF Samwha caps sound better than the 1000uF ELNA caps (to my ears / by quite a margin) the 3,3uF bypass caps (to my ears) are also a lot better than the 100nF stock bypass caps.

YMMV.

[PinkFloyd]

Mar 21, 2010 21:00:10 GMT @heliharris said:

These "tails" make it a LOT easier to adjust the bias.... two multimeters is also the best way to go (with crocodile clip probes) you can set both channels up with ease and don't have to poke about.... just clamp the tails and clamp the other test lead onto the outside of the phono sockets..... all you have to do now is turn the trimpots until both multimeters show 13V...... re-adjust after one hour.

Made up extension leads as per Mikes instruction to day and have since checked The bias on mine. 7.1v left 9.2v. right with standard valve, way below optimum of 13v, it sounds a lot nicer now to my ears. Would have thought Indeed would have taken more care in at least setting the first 10 properly(no one so far has found there`s to be correct). Its a good job i don`t send my work out like that, if i did, i wouldn`t get any more to do!!.

Yup,

Pretty poor and it is evidently clear that the trimpots are simply set to "approximate" positions at the factory. As I have said before.... the amp costs £59, they will not have an employee sitting there for an hour farting about with the bias (on each amp)..... the sample model came with a little sticker on it which said "6922 matched bias" so that one was obviously set by hand.

I don't mind having to adjust the bias myself (I actually enjoy it) but customers should be told how to do it AND told that they will NEED to do it before use.

Seems "we" are doing it for them...... again, I don't mind.... I enjoy pottering about with things BUT the manufacturer should be more specific with their instructions.

Mick the amp will sound OK at the settings you mention but far from optimal..... I'd like you to try 12V - 17V see what you prefer. There is obviously no "critical" rule and whatever suits your ears

Mike.

[PinkFloyd]

Peter has replied re: the 85C capacitors parked up right next to the heatsinks and I DO agree with him re: the capacitor working better when it is warm:

" Hi,
I willfully to install the output 85C Elna capacitor near heat sink ,it get a (heat effect); the ESR and recharge time of the capacitor will be better when the temperature increase, the working of capacitor is a chemical reaction, it has most active in hot condition just like the car battery also good in summer but bad in winter.
We have experiment that 85C is better than 105C when there work in same temperature (-4C-80C range)
Don¡¯t worry about the life of capacitor; it can work over ten year on G2.
Peter"

I totally agree with this Peter, BUT a 105C cap will last a lot longer than an 85C type and a 135C type will last even longer...... they are all subjected to the same heat in the G2 so will all "fizz" just the same (chemical reaction)...... there's obviously a LOT more to it than this, which is what makes audio so interesting (maybe Frans can comment on optimum capacitor temperatures, or maybe Leo)...... to my "ears" the 2200uF sound better in this position and the 135C rating is an added bonus as far as longevity is concerned.

Your point regarding an "optimum" temperature for a capacitor to do what it does best (audibly) is a FACT, IMO... they do seem to sound better the warmer they get..... this is good, short term, with 85C rated caps (how do you know they will last ten years in the G2?) but "heat", unfortunately, shortens their "in spec" operational life. I'm sure Frans will present you with the Mathematics ;D

I'm a "belt and braces" type of guy Peter (this means I use everything at my disposal, however seemingly unnecessary) to ensure that reliability is "guaranteed"...... There's NOTHING wrong with the G2 "as is" and, being an open frame design, the 85C caps will probably last for many years.... plenty of air circulating around them......

Don't worry what "I" have to say, I'm just another mouthpiece in an ocean of gobshites....

Mike.

[Deleted]

Mick the amp will sound OK at the settings you mention but far from optimal..... I'd like you to try 12V - 17V see what you prefer. There is obviously no "critical" rule and whatever suits your ears 8-)Mike.

Mike, as you and others on here know, i have no real understanding of the science involved in creating a good , nice sounding Amp, i only know what enters my ears, and how it sounds to me.

What i do know is, that after adjustment to 13v bias on both L/R channel there was, a positive improvement in SQ.

My point is, most of these Amps will be going to customers, who have no interest in Forums, bias settings, etc they just want to plug in and go listen to the music, so its in Indeed`s interest to make sure that these Amps deliver the best possible SQ out of the box, for Mr Joe Blogs, the plug and play man.

At Least get L/R settings the same, be 12, 13, 17, i can do it in 10mins (5mins,1hr later another 5mins) job done, and sounding much better.

Anyhow, once i get used to the "new" sound of my Amp, i will give the other voltage recommendations a listen in 1v increments Mike, and report back. Thanks.

Mick.

[clausdk]

I do not know what the difference is in those caps, but if it is isolation, the core of the more heatresistant types will not get as hot as the lower type, therefore not work as fast, I do know my chemistry

Speed would be the only difference, so if a bigger cap is used it might level things out a bit, but I am talking without knowledge (as usual )

just a thougth

[MaN227]

Mar 21, 2010 23:36:00 GMT @heliharris said:

My point is, most of these Amps will be going to customers, who have no interest in Forums, bias settings, etc they just want to plug in and go listen to the music, so its in Indeed`s interest to make sure that these Amps deliver the best possible SQ out of the box, for Mr Joe Blogs, the plug and play man.

that is exactly the point I was trying to make in the indeed/bravo thread for bias. I bring it up and its like you can't expect them to take so much time to do that for each amp. no correct me if I'm wrong a fair amount of time goes into assembly of EVERY amp a few minutes to amoung the hours does not seem like to much to ask. and the main point it why on earth would ANYONE sell an amp that did not sound the absolute best it can sound from the first time you turn it on. as I stated my was SUPER out of whack. 19 and 11 v .
it seems money is the motivating factor pumping them out. now if you have a business selling ANY item and you really do want many dollars a few extra minutes on each amp amoung the hours spent seem what one would want to do . no one gave me any GOOD reason as to why before and i don't suspect I will get a good one now.
quality control has to play a bigger part than getting them into the mailing box 5 minutes faster. unless you would like your products to be a hit or miss. many many will buy this amp and it be way out of whack and what? this all really makes me wonder. you don't see clear concise directions from maker why? well for one it shows that they did not take the time themselves to ship out the best product, in all regard , that they can. sad.
think about being someone who does not visit hifi forums and you buy this out of whack amp you would have no clue and or reason to suspect its not as good as it can be. nor would you have any idea on adjust ing bais or that one can even do it.
everyone is nice that is selling you something.

it seems many here don't see things as I do, its cool. but tell me how happy any of you would be about any other amp that you spend your good money on is not even adjusted to its best or at least a standard setting.
it only makes your product better, better product , happier customers , more sales. bigger BANK. but what do i know.
Imagine a nice wooden dresser you just spent good money on and I who made it and sold it to you. I cut every piece precisely, assemble it to an exacting standard and chose to only sand half of the dresser smooth and just applied the stain and clear over it . one side smooth like a babies bum and the other is very rough to the touch.
make sense? .....yeah that's what I thought.

[attilio]

think about being someone who does not visit hifi forums and you buy this out of whack amp you would have no clue and or reason to suspect its not as good as it can be. nor would you have any idea on adjust ing bais or that one can even do it.everyone is nice that is selling you something.

You make a very good point, Indeed should provide instructions on bias setting at the very lest. After all how much time would that take? However if they did that they would have to redesign the amp so the user can do the adjustments easily, this I suspect would add to the final cost of the amp. So do we want a more user friendly amp at higher cost? Or an enthusiasts amp that we enjoy tweaking.


attilio

[attilio]

Also how many members of the public have a multimeter ?
This is an enthusiasts product at this stage of it's development.

attilio

[Deleted]

Just did some initial testing on this amp with 33 and 300 Ohm dummy loads attached (to mimic HP's for measuring no other reason)

It is difficult to explain this to technical noobs but there are some problems with these amps.
All related to the bias and glowing voltage.
When the LM317's are NOT biassed properly (so bias must be set above 15 V with NO music input attached) there is SEVERE and I mean SEVERE crosstalk between left and right channel that is very unlineair meaning one half of the sine wave has huge crosstalk and the other half just a little and linear.
This is SO measurable that it must be audible too.

pics would be needed to make this visible and am thinking of writing up a report if people are interested.

max output voltage @ 300 Ohm = 3.9V RMS (above 1.9V RMS the distortion becomes clearly visible)
max output voltage @32 Ohms = 1.6V RMS (above 1V RMS the distortion is very visible)

Both with HUGE asymmetrical L to R crosstalk distortion above a certain point.

check for yourself if your amps have this too please (general call to owners) in the following manner.

put a music signal on the left OR right channel only (leave one plug out or preferably short that input with a plug that has the middle pin and ground connected).
cranck it up and play music on that one channel.
Listen to the left cup (get the right one from your ear so you don't hear it) and then listen to the other channel alone (lift the other cup so you don't hear it)

Tell me what you HEAR... (you SHOULD hear absolutely nothing when no signal is attached to one input on the channel that is NOT connected)

the higher the volume the louder the distortion above a certain point.

Be very carefull with DC coupled preamps connected to this amp b.t.w.
A DC component in the signal path will throw the bias off.

When you bias below 15 volts the bias voltage will vary greatly with music signals (volume) applied. (it will run up depending on the music amplitude)

In Technical terms the amp/design is suspect.
But it's all about the music is it.
So if you like it... just use it and enjoy it and forget about the technical shortcomings of the design.

please check the one channel thing someone and report back to me or in the thread.

The effect will very likely be masked or not even there when both channels have a signal as the bass usually is mono anyway.

Now about the caps close to the hot parts.. ;D
If this was deliberate (putting them close to the LM317's) then WHY is one channel very close and the other on a greater distance ... ?
Makes no sense to me at all and appears to me as a bogus reason.
If this were the reason they would have equal distances and be thermically coupled.

To my knowledge a higher temp of the caps is NOT benificiery at all.
Of coarse in audioland different rules apply and makes the sound 'warmer'.

I will continue research on it and IF I decide the amp is worthwhile doing the autobias mod I will go ahead with it.

a puzzled guy...

please do the 'one channel test' someone.

<edit>

I now understand the vol pot problems as described by Ian.
The vol pot used is a linear one... not a logarithmic one as one might expect as a vol pot in an amp..
Very weird log pots aren't that more expensive.

easy fix:

Put a 2k2 (2200 Ohm) resistor between the center contact of the vol pot and ground.
Do this for both channels and the vol pot will behave a bit more like a logarithmic vol pot ...

[attilio]

This may be of interest, www.slugsite.com/archives/1289

[MaN227]

I'm liking the G2. hiss buzz and such is less, still there but less than the G1. seems to me that added heatsink does indeed help out, there is plenty of heat in it fo sho not quite AS hot as the front most ones but hot all the same.
HP jack i find miles better than on the G1. I like the sound it delivers with stock 6922 tube.

which leads me to a question.... does the amount of v bias have any effect on the noise floor? if not what actually does I hate that hiss. the x-can v3 is much lower noise floor. damn near black. but I suppose that is unfair to compare at the price diff.

I guess the question we need to ask is at its price point is there better? I've not heard better for the money.

now I have to visit the skytronic's thread and get these things out from behind their heavy velvet curtains.

is anyone else's red led much much more dim as mine is? I know its just trivial, just curious. I thought I liked it better but I don't i like the better reflected/refracted light that the G1 gives better than the G2.

the wee metal feed look snug to the plexi yet my teeters to and fro bummer I will find some feet at the hardware store as I don't like the tipping or sliding it does. I need a rotozip or similar I'm ready to beef this thing up a bit in weight.

I very much like Micks clear feet, after all it works way better with the lights.
but bulky metal is what I'd love to have. that's another story.

[attilio]

Thankfully I have not ordered mine, I will give this a miss.

attilio

[Deleted]

All technical deficiencies aside, many of the guys here love their little amplifiers for the fun and novelty value, as well as being a good listen, and at a very reasonable asking price.
They will also be the first time that many have experienced a valve amplifier.
I feel sure that if QC had been better, we may not have been digging so deeply into their major design inadequacies
I do not feel that it is helpful to continue further into what is wrong with them, but to concentrate on showing how they may be further improved economically , without losing their character.
I look forward to further photographic posts of Mike's mods, as well as what Frans can come up with.

SandyK (Alex)

[Deleted]

The fun and novelty part is very true.
The sound won't be bad either otherwise it wouldn't have that many fans.

On the testbench it is horrific.
All other amps I measured did much better on the testbench.

Alex has a point though...

Not bashing the thing just trying to find out if the bad measurements can explain the sound that is perceived as better than amps that perform well on the testbench.

now busy at the temperature household regarding Peters remark about the deliberate placement of the output caps near the LM317's.

the generated temp from the LM317 with a higher bias voltage will be MUCH higher then when a low bias setting is used..
Also contradicts the temp/bias story in itself as a higher temp (higher bias voltage) would make the caps sound sweeter ?