Mystery revealed: UpTone Audio "UltraCap™ LPS-1"

[Jamesroy]

Does any of this matter to us who use a Mac Book Pro?

[Jud]

Does any of this matter to us who use a Mac Book Pro?

 

Not if you are looking for a power source for the MacBook Pro.

[zegel]

Alex, what made you to decide to order 150 boards? Are you confident this amount matches the demand? Reason for asking I am a bit nervous not being part of bach 1 and to have to wait longer.... ( Am I correct batch 2 will take 40 + 40 days from now minimum?)

[Superdad]

Alex, what made you to decide to order 150 boards? Are you confident this amount matches the demand? Reason for asking I am a bit nervous not being part of bach 1 and to have to wait longer.... ( Am I correct batch 2 will take 40 + 40 days from now minimum?)

 

Don't worry. Everything will be ready and shipping a lot sooner than 40 days. The first 150 boards are going to be here before the end of September, and another 100 can come about 18 days from whenever it seems appropriate (i.e. If preorders for the first ones are actually rockin'). We already have 250 cases, etc. here.

 

Right now we are just trying to get ready for assembly/programming/testing/packaging. Plus John is finishing up code for it, I'm setting up a Linux machine for production FPGA flashing, and working on the web info/order page.

 

At the same time, I still have subcontractors here daily as we are more than 3/4ths done with the new production facility addition. Picking out paint, flooring, door and window moldings while the HVAC and last of the electric and plumbing go in. Lots of noise today as the place is on a grante rock and the plumber is drilling and jackhammering to get a 4 cubic foot sewer ejector pump/barrel into the ground. He is threatening to dynamite!

 

I stupidly wasted yesterday and night with migrating myself to a lovely 5K Retina iMac (bought last year for myself but just used by my bookkeeper). I decided to finally do it now because all my Adobe licensed s/w has all been on the new machine for some time so I have not been able to run Photoshop, Indesign, etc. at my desk. The process went well--except for the e-mail migration which I am still fighting with (Apple made some big changes with El Capitan which have caused mail transfer issues for a lot of people.). E-mail is where I spend my days, so I'd best get this fixed in the morning. Argh.

[Superdad]

How do I get myself on your mailing list? BTW I'm an existing customer and have ordered USB Regen. Does that qualify to be on mailing list?

 

As a past customer you are automatically on our mailing list.

 

And at the bottom-right of our homepage there is a place for anyone to enter their e-mail address to be add to our private mailing list (don't worry, I never give/sell the list, and e-mail missives from us come extremely infrequently).

[Pepsican]

Right now we are just trying to get ready for assembly/programming/testing/packaging. Plus John is finishing up code for it, I'm setting up a Linux machine for production FPGA flashing, and working on the web info/order page.

 

Sorry if this has been asked before, but does that mean there will be firmware updates in the future (if that is even possible)?

[JohnSwenson]

Sorry if this has been asked before, but does that mean there will be firmware updates in the future (if that is even possible)?

 

There is a micro SD card slot on the board, which is there to field upgrade the firmware if needed. We are not PLANNING any firmware updates, but the mechanism exists if it becomes necessary.

 

John S.

[rickca]

Vinnie Rossi's upcoming mini PURE-DC-4EVR uses a built-in linear regulated charging circuit with no SMPS or DC-DC converter circuit. This sounds really slick. Can you please explain why you didn't include something like this in your design of the LPS-1? Is it just to keep the price of the LPS-1 reasonable or are there other considerations?

[firedog]

Vinnie Rossi's upcoming mini PURE-DC-4EVR uses a built-in linear regulated charging circuit with no SMPS or DC-DC converter circuit. This sounds really slick. Can you please explain why you didn't include something like this in your design of the LPS-1? Is it just to keep the price of the LPS-1 reasonable or are there other considerations?

 

I think you are looking at $399 vs several $thousand. We will see when Vinnie's device is actually produced.

[rickca]

@firedog it's $395 for LPS-1 vs $795 pre-order for the MINI PURE-DC-4EVR. So you're right, it's a big difference in price but not as big as you suggest. Vinnie's MEGA unit will be in the price range you mention.

 

Unless John and Alex want to talk about this built-in linear regulated charging circuit idea, I won't post anything further about Vinnie's product in this thread. I'm excited about the LPS-1 and I don't want to sidetrack the discussion.

[lmitche]

Nicely said!

[JohnSwenson]

Vinnie Rossi's upcoming mini PURE-DC-4EVR uses a built-in linear regulated charging circuit with no SMPS or DC-DC converter circuit. This sounds really slick. Can you please explain why you didn't include something like this in your design of the LPS-1? Is it just to keep the price of the LPS-1 reasonable or are there other considerations?

 

There are several reasons for not using a linear charger:

 

1) the charging voltage has to be greater than the maximum voltage on the capacitor string. To do this with a linear charging circuit would take a minimum of 15v input, severely limiting the choice for feeder supply. Alex really wanted people to be able to use the 7.5V supply that came with the REGEN, that would be impossible with a linear charging circuit.

 

2) the linear charging circuit would dissipate a LOT of power, way more than the current design, this would require a larger more expensive case to get rid of all that extra heat.

 

3) the parts to implement it would cost more.

 

I actually built the first prototype with a linear charger, boy did that get HOT!

 

From a sound aspect there is no reason for going with a linear charger, remember the output is isolated from the charging circuit. The only way any of the noise from the charging circuit can get to the output is through the transistors that switch the bank from charge to output. That can only happen through the capacitance of the transistors, so I choose very low capacitance transistors. They are hard to drive so it took a lot of work in the design, but they are working great now.

 

It is possible for the switching circuits to emit EMI from the board (they do in fact do this), but I spent a lot of work on the board layout to minimize this. In addition the board is well shielded in the case which blocks the fairly low EMI coming off the board. With my most sensitive instruments I could not detect any EMI coming off the assembled system.

 

The other possible mechanism is that some noise from the switching charging circuit could get "out" of the box through the input "feeder" cable, I have measured this and see no difference in noise on the DC coming from the feeder supply. At the maximum sensitivity of my equipment there is definitely some noise already there, no matter what feeder I use. With the LPS-1 doing it's thing charging away I cannot see any change to the voltage from the feeder. So it looks like the measures I took in the LPS-1 to block any switching noise from getting back on the feeder line are effective.

 

John S.

[Jud]

@JohnSwenson : why not just use an old fashioned linear unregulated charging supply? Is there any reason the charging voltage needs to be precise?

 

 

Room treatments for headphone users

 

Thought I remembered John mentioning something about this early in the thread, but don't recall what it was (if it actually happened).

[JohnSwenson]

@JohnSwenson : why not just use an old fashioned linear unregulated charging supply? Is there any reason the charging voltage needs to be precise?

 

 

Room treatments for headphone users

 

Are you referring to a feeder supply for the existing circuit?

 

If so the problem with an unregulated supply is that the voltage varies radically from full load to no load. The way the LPS-1 works causes this to happen all the time, it charges at full current for several seconds, then turns off almost completely, then turns back on full bore again etc. This would be causing large voltage swings with an unregulated feeder supply. This could easily produce some pretty large spikes that could eventually fry things on the input circuit.

 

John S.

[romaz]

John, the LIO's PSU has a reported output impedance of about 18 milliohms (although it's not clear throughout what frequency range). Can you comment on the output impedance of the LPS-1?

[jabbr]

Are you referring to a feeder supply for the existing circuit?

 

If so the problem with an unregulated supply is that the voltage varies radically from full load to no load. The way the LPS-1 works causes this to happen all the time, it charges at full current for several seconds, then turns off almost completely, then turns back on full bore again etc. This would be causing large voltage swings with an unregulated feeder supply. This could easily produce some pretty large spikes that could eventually fry things on the input circuit.

 

John S.

 

Interesting. I was assuming that each supercap is being fed alternately so the "duty cycle" for current would be more like ~80% and capacitance (feeder), either straight or multiplied, could smooth out.

 

A good fast cap multiplier might smooth the voltage out without wasting too much power. Just a thought.

[zilch0md]

There are several reasons for not using a linear charger:

 

[snip]

 

From a sound aspect there is no reason for going with a linear charger, remember the output is isolated from the charging circuit. The only way any of the noise from the charging circuit can get to the output is through the transistors that switch the bank from charge to output. That can only happen through the capacitance of the transistors, so I choose very low capacitance transistors. They are hard to drive so it took a lot of work in the design, but they are working great now.

 

It is possible for the switching circuits to emit EMI from the board (they do in fact do this), but I spent a lot of work on the board layout to minimize this. In addition the board is well shielded in the case which blocks the fairly low EMI coming off the board. With my most sensitive instruments I could not detect any EMI coming off the assembled system.

 

The other possible mechanism is that some noise from the switching charging circuit could get "out" of the box through the input "feeder" cable, I have measured this and see no difference in noise on the DC coming from the feeder supply. At the maximum sensitivity of my equipment there is definitely some noise already there, no matter what feeder I use. With the LPS-1 doing it's thing charging away I cannot see any change to the voltage from the feeder. So it looks like the measures I took in the LPS-1 to block any switching noise from getting back on the feeder line are effective.

 

John S.

 

I very much appreciate this explanation, John. Thanks!

 

Please jump in and correct me if I say something incorrect here, but in short, the power coming out of the LPS1 caps is absolutely devoid of any noise that came in on the energizing supply - we can all take comfort in the isolation provided by this design - but it's perfectly reasonable, in my opinion, for us all to accept the fact that nearly insignificant levels of noise can come from various unavoidable sources on the LPS-1 board itself - all of which have been suppressed to the best of your ability.

 

This is analogous to your honesty regarding the fact that the USB Regen can itself generate small amounts of noise. As a community, there are some who seem to be so noise-phobic that any attempt to acknowledge or quantify noise of any kind that emanates from your designs is met with borderline hysteria.

 

Keep hitting us with the truth. It's refreshing. :-)

[Superdad]

Gents:

 

While we have been very forthcoming about most of the details of our UltraCap LPS-1, I would appreciate avoiding turning this thread into a deep technical discussion of the merits and limitations of the market's only two bank-switching, microprocessor-controlled linear power supplies. Neither ours nor the just-announced Vinnie Rossi units are shipping yet. There are aspects which are similar and which are different. Let's debate the merits and performance once the products are released and in people's hands.

 

John does have the intention to produce interesting measurements for all to see, but right now he is focused on finishing up the programming and other details. Besides, as he explains it to me, there does not exist a piece of test equipment on the planet that makes it easy to measure and plot power supply output impedance versus frequency. He is designing his own set up to do so, but that takes lower priority than other things.

 

Thanks,

--Alex C.

[jelt2359]

Could I use this to power my FMC optical units?

[MikeyFresh]

Could I use this to power my FMC optical units?

 

 

Read Post # 9, then compare the LPS-1 specs with the requirements of whatever you intend it to power.

[Superdad]

Could I use this to power my FMC optical units?

 

Actually, assuming your FMC can run from one of the UltraCap LPS-1 voltages (user switch selectable, 3.3V, 5V, or 7V), the DAC-end of an FMC pair would be a fantastic application for our new supply. That is because it will preserve the galvanic isolation you get from the fiber connection, and unlike any other PS you might use (aside from batteries) not allow any system leakage currents to get back in at that point in your chain.

[sonata]

Actually, assuming your FMC can run from one of the UltraCap LPS-1 voltages (user switch selectable, 3.3V, 5V, or 7V), the DAC-end of an FMC pair would be a fantastic application for our new supply.

Alex, I have the TP-Link MC100CM - its power requirements are 9V/0.6A or 5V/1A. By my reckoning, it should draw about 0.8A at 7V.

 

So can you confirm that the LPS-1 (energised by the Meanwell PSU from the Regen), would be OK to power the DAC end of the FMC plus the microRendu which draws about 0.2A at 7V?

 

Regards, Rob

[JohnSwenson]

Alex, I have the TP-Link MC100CM - its power requirements are 9V/0.6A or 5V/1A. By my reckoning, it should draw about 0.8A at 7V.

 

So can you confirm that the LPS-1 (energised by the Meanwell PSU from the Regen), would be OK to power the DAC end of the FMC plus the microRendu which draws about 0.2A at 7V?

 

Regards, Rob

 

If the FMC really does draw 08A then you will not be able to drive both from the same LPS-1. The .2A for the microRendu is kind of averaged during normal running time, I mentioned that for determining power usage for thermal issues. Actual peak current can be twice that (0.4A) which will not work with .8A from the FMC.

 

Actually using the same LPS-1 for both the FMC and a microRendu is not a great idea anyway. The MicroRendu DOES have galvanic isolation between the Ethernet input and the USB output, using the same LPS-1 to drive both shorts out that isolation. You will probably be better off using the LPS-1 on the microRendu and something else for the FMC.

 

John S.

[adamaley]

Will it be advisable to use one output of the JS-2 power supply with a Y-splitter for the two FMCs, then use the other output to energize the LPS-1 to the Microrendu? This is my planned setup for once the LPS-1 becomes available. For now it will be JS-2 directly to the Microrendu, then the second output split to the two FMCs.

[sonata]

Duly noted. Thank you John for that.