Little Dot MK8SE / MK6 Super Mods (All verified mods are on first page)

[CopperFox]

Going to look around at latest films for something suitable.

There are some things I learned while looking at those that may be useful here. The film caps with these capacitances come in some different dielectric materials and they have clear differences between them in thermal-electric properties which sets them apart from each other. Polypropylene caps are the best and the rest have higher temperature dissipation which causes both higher signal loss and also possible further degradation from internal overheating. Polypropylene caps are also slightly larger than polyethylene or polyester caps.


Here's a list of some current caps in the ~100uf mark, most of which are currently availabe. Sorted by dimensions from largest to smallest.





The polypropylene caps should have no problems with heat or varying levels of AC signal.

For some of the caps the manufacturers have specified limits as to how much AC voltage they can handle without overheating (which is defined as higher than 10C temperature rise from internal heating). As an example, here's a clip from the TDK B3252X series data sheet:



Here the 63V version is specced as being able to handle AC signals up to 10kHz - with declining voltage - but the 100V version (which I didn't find in stock anywhere) looks like it can do two volts at 20kHz.

And here is the similar graph(s) from the Vishay MKT1820 data sheet:



The graph for the 63V version lacks the 100uf capacitance but it looks like it would slot in somewhere slightly below two volts at 20kHz. The graph for the 100V version is not much better and neither is the 250V version.

The film caps with appropriate specifications here are designed for DC linkage and not high frequency AC signal passthrough as that will apparently cause more heat to be dissipated by the capacitor. The polyproplene caps however are mostly immune to this as their dissipation factor (and therefore internal heating) is an order of magnitude lower than the polyethylene or polyester caps (<0,1% vs <1% vs <1,5%). The lower dissipation factor by itself also will cause them to have better sound in audio use. Conversely, the non-polypropylene caps' sound may be actually be worse than some electrolytics at least in some respects/aspects (They might result for example in a sound that is smooth but otherwise lacking).

So from the available options the 100uf or 90uf 400V DCL Wima would seem the option of choice taking into account dimensions and thermal-electric properties. The 90uf is not stocked anywhere though and the 100uf is slightly larger. By the Wimas' spec sheet, the 400V 100uf has the exact same dimensions as the 800V 40uf version which has been observed to fit inside the standard case a few pages back here . So that could be doable.

In any case, the main point here is that polypropylene caps should be far better than the other film options in audio use so they should be used here if at all possible despite their slightly larger size.

(After having ordered the 560uf Vishay film caps which also were from the MKT1820 series, I noticed that the graph in the data sheet shows they should be able to do about 0.2V at 20kHz without overheating. While having them in the amp I did notice there was at least slightly more residual heat in them than the other components an hour after switching the amp off. I'm not sure if the amount and duration of heat from my use would be enough to make a difference in performance as the data sheet also defines a "maximum operating temperature for limited time" as 200 hours at 150C at 30% of rated DC voltage. However, I did also notice that the Wimas were clearly better sounding right from the start.)

 

[CopperFox]

I believe this is due to the phase effects of a slower charge/discharge cycle, it is probable that the signal is affected by the slower cycle whereas a fast cycle will allow transients to occur naturally faster with less problems of clipping.

I had been thinking inductance (the tendency to resist changes in voltage due to magnetic effects caused by the component's structure) may have an effect there as well. It sounds like something which may cause a burst-like or jagged characted in audio use.

 

[baronbeehive]

In any case, the main point here is that polypropylene caps should be far better than the other film options in audio use so they should be used here if at all possible despite their slightly larger size.

Yes, I was going to bring up this point about film construction, it seems there is no agreement about which material is best but this is useful information here.

However, I did also notice that the Wimas were clearly better sounding right from the start.)

Good information, I think the Wimas are also the smallest size, anyway it's good to know what you've found, I'm already thinking along these lines.

The one thing I've drawn a blank on so far is the ratio of lytic to film capacitance but as you say 100uF would seem to satisfy all conditions whatever the ratio. I was thinking along the lines of a compromise of 150uF just in case the ratio is right at 2:1. But from what I can find out, together with your findings with your caps it seems that almost any size would work although it would not be good to go too high as various factors could affect sound, not to mention size.

The point about whether it would fit in the chassis is irrelevant to me as my chassis is choc full anyway, so I would have to extend it .

Edit: It would seem that film caps would be better in practically any use over lytics, however lytics are good at storage so that points to lytics being OK for PSU functions so maybe film would not be necessary here. A further point is that film caps might have problems dealing with the ripple through the PSU, this would need further investigation to confirm, which might make them not ideal for PSU functions which you alluded to by the post above where you said film caps were designed for DC not AC, although I'm sure you can still use them for PSU.

One more point worth mentioning is that there can be interference betweem caps if placed to near to each other, so maybe bear that in mind when trying to squeeze them into a crowded chassis!

 

[baronbeehive]

Going to look around at latest films for something suitable.

I've tried to find the blue Wimas but it seems the FKP1 only go up to 4.7uF, at Mouser at least.

 

[baronbeehive]

I had been thinking inductance (the tendency to resist changes in voltage due to magnetic effects caused by the component's structure) may have an effect there as well. It sounds like something which may cause a burst-like or jagged characted in audio use.

Yes, also their behaviour is not linear unlike film caps, and also they can be rather unstable in output under some conditions.

 

[Maxx134]

Voltage :
Not much voltage here, the "original" 50V is more than enough

Last time I measured (years ago) I found it lower than half that rating.

The vu meters were beginning to show imbalance with one channel rising in output and the other decreasing

Them meters will show according to the bias of the output tubes. Sometimes old tubes will make it vary.

so it wasn't the load level in itself causing that

Yep I noticed lower impedance or planar headphones will affect the meters a bit more also (White Cathode follower design).

Here's a list of some current caps in the ~100uf mark, most of which are currently availabe. Sorted by dimensions from largest to smallest

Nice list.

I'm not sure if the amount and duration of heat from my use would be enough to make a difference in performance

I am so very impressed by all this research and this is the type of depth that actual dedicated designers do when planning.

That being said, I like to add some more observations which help in determining parameters.
Being that the caps will be on the underside of the circuit board, and below the amp, I do not forsee the tube ambient temp (around 100C avg) affecting the ambient temperature below its socket too much.

I do not believe the ambient temperature of the caps on the board to rise above 40C with all the resistors & PSU heat combined in there.

Also, the voltage drop measured across the Cathode cap is actually fairly low, so I believe we can also go for lower ratings.
The test would be to observe the performance difference "audibly", and also by checking if the bias meters dip when bass peaks hit(recovery)...

I thinking to try doing one side (not both) and then compare the other side any differences, with a monural source.

 

[baronbeehive]

According to Mouser their DC Link are all Polypropylene whereas their other capacitor types are Polyester, Polyetheline and Polypheneline. Not sure they have all the sizes atm, only the MKP6 go up to and beyond 215uF so might need to shop around.

Edit: On the wima website the MKP6 and MKP4 DCL go up to and beyond 200uF.

NB: I just reread page 1 and I see that the ratios of lytic to film are from 2:1 up to 10:1 according to Redge, so either I was right or we were both reading from the same... hopefully correct sources!

 

[CopperFox]

Yes, I was going to bring up this point about film construction, it seems there is no agreement about which material is best but this is useful information here.

Good information, I think the Wimas are also the smallest size, anyway it's good to know what you've found, I'm already thinking along these lines.

The one thing I've drawn a blank on so far is the ratio of lytic to film capacitance but as you say 100uF would seem to satisfy all conditions whatever the ratio. I was thinking along the lines of a compromise of 150uF just in case the ratio is right at 2:1. But from what I can find out, together with your findings with your caps it seems that almost any size would work although it would not be good to go too high as various factors could affect sound, not to mention size.

The point about whether it would fit in the chassis is irrelevant to me as my chassis is choc full anyway, so I would have to extend it .

Edit: It would seem that film caps would be better in practically any use over lytics, however lytics are good at storage so that points to lytics being OK for PSU functions so maybe film would not be necessary here. A further point is that film caps might have problems dealing with the ripple through the PSU, this would need further investigation to confirm, which might make them not ideal for PSU functions which you alluded to by the post above where you said film caps were designed for DC not AC, although I'm sure you can still use them for PSU.

One more point worth mentioning is that there can be interference betweem caps if placed to near to each other, so maybe bear that in mind when trying to squeeze them into a crowded chassis!

Yes the DCL Wimas seem to have the most capacitance for size in the poplypropylene caps. (edit: removed mistaken quote of current ratings)

According to Mouser their DC Link are all Polypropylene whereas their other capacitor types are Polyester, Polyetheline and Polypheneline. Not sure they have all the sizes atm, only the MKP6 go up to and beyond 215uF so might need to shop around.

The 400uf 400V seems to be out of stock but can be backordered also in small quantities at mouser. The 200uf 400V can be found in stock at tme.eu where I got mine from.
Two of those in parallel would also be 400uf but take more space. Perhaps there might be some other brands that could fit in an extended case as well? Haven't yet looked at that.

 

[Whitigir]

So what amp is to buy in order to have all this fun? There are non on Amazon

 

[CopperFox]

So what amp is to buy in order to have all this fun? There are non on Amazon

littledotus.com is the US/EU distributor and it's also available at shenzenaudio.com and from various sellers at Aliexpress. 11/11 is coming up in two and a half weeks and there will probably be discounts from the Chinese sellers so that might be a chance to save some cash.

The black colour variant has not been available for the first half of the year, don't know about the current situation on that. Silver is the other option and that has been steadily available.

 

[Whitigir]

This is OTL correct ?

 

[baronbeehive]

Here is an online tool I've found that will show ideal values for the cathode bypass cap which should take the sweat out of decision making on the size of cap: https://www.ampbooks.com/mobile/amplifier-calculators/cathode-capacitor/calculator/

So whether to use 100uF or 200uF should become obvious and we might be able to dispense with guesswork on ratios. However it doesn't deal with the differences between lytics and film caps but it should show the point at which it is overkill in going any higher in values of microfarads for example. For example it would show up on the graph whether there would be anything to gain having reached an optimum point, on going any higher in value because the graph would stay much the same if that makes sense. It would also show optimum values for bass, mids and treble so depending on your preferences you could see instantly what cap to use.

 

[baronbeehive]

This is OTL correct ?

OTL and OCL.

Edit: Be careful who you buy it from, littledotus has said that they take care in testing thoroughly before sending out and also offer better redress, but you could check the other LD thread where this has been discussed recently due to some bad experiences that some have had:https://www.head-fi.org/threads/little-dot-mk-vi-little-dot-mk-viii-se-owners-unite.522099/page-181

 

[baronbeehive]

The 400uf 400V seems to be out of stock but can be backordered also in small quantities at mouser. The 200uf 400V can be found in stock at tme.eu where I got mine from.
Two of those in parallel would also be 400uf but take more space. Perhaps there might be some other brands that could fit in an extended case as well? Haven't yet looked at that.

Sounds good! I know that member coin, used 470uF cathode film caps without any problems and raved about them if I remember right! But there is some doubt about going too high, the higher the value the more bass and the less treble, and visa versa.

 

[baronbeehive]

Yes the DCL Wimas seem to have the most capacitance for size in the poplypropylene caps. The RMS ripple current rating for the 400uf 400V is 43A in the datasheet which probably should be enough?

That should be way over the top on a 170 odd volt B+ but I will get back to you on that.