Announcing the next generation: UltraCap LPS-1.2 (Now 5/7/9/12V! ; Now with parallel LT3045 regulators!)

Runner · January 24, 2018

Hello guys!

Can I use your UltraCap ™ LPS-1.2 to feed the CHORD QUTEST DAC?
In the description of UltraCap ™ LPS-1.2 you write:
Maximum guaranteed output current-at all voltages-is still 1.1A.
The CHORD site shows the power consumption of the CHORD QUTEST
5v 2amp Micro USB

Regards, Yuri

pl_svn · January 24, 2018

Rob Watts wrote (Qutest's official thread on Head-Fi) that a 5v/1A power supply is fine

bobfa · January 24, 2018

Yurt,

I am not pretending to answer for Alex but a couple of observations: The QUTEST appears to only have a MicroUSB input. The specs say that the “power supply” is rated at 5V 2A. It does not say what the device draws. It also has not shipped yet so we really do not know what the power requirements of the device is!

I think that using a Micro USB connection for power for an about $1200 DAC seems; well odd to me.

ElviaCaprice · January 24, 2018

1 hour ago, bobfa said:

Yurt,

I am not pretending to answer for Alex but a couple of observations: The QUTEST appears to only have a MicroUSB input. The specs say that the “power supply” is rated at 5V 2A. It does not say what the device draws. It also has not shipped yet so we really do not know what the power requirements of the device is!

I think that using a Micro USB connection for power for an about $1200 DAC seems; well odd to me.

Yes, we do, Rob Watts is the designer and has given out this information over at Head-fi. 5V 1A is more than sufficient.

Superdad · January 24, 2018

3 hours ago, pl_svn said:

Rob Watts wrote (Qutest's official thread on Head-Fi) that a 5v/1A power supply is fine

Yep, and all you need is one of these:

https://www.amazon.com/gp/product/B01G6EBGWO

I know at least 10 people have already ordered an UltraCap LPS.1.2 for use with their Qutest. Perhaps they will receive both devices at about the same time?

==========

By the way gents: This is shaping up to a be a huge launch. Pre-orders for the LPS-1.2 are through the roof. I hope the cases and boards arrive really soon so we can start assembling.

THANKS EVERYONE!!

--Alex C.

[email protected] · January 25, 2018

1 hour ago, R1200CL said:

Is adding a external LT3045 box for the LPS-1 something you would consider as a possible future product for us LPS-1 owners. ?

I am planning to have a case for my LT3045 Board with 2.1 x 5.5 mm dc power jack (as input) and short cable, soldered to regulator board with 2.1 x 5.5 mm Plug.

austinpop · January 25, 2018

Señores @Superdad y @JohnSwenson

Simple question for you. Given the LPS-1.2 can now handle a larger voltage range, from 5-12V, will it sound better at one or other end of the range? I ask this because I wonder if lower voltages incur more stages of regulation?

Looking forward to receiving my units to find out for myself!

Superdad · January 25, 2018

5 minutes ago, austinpop said:

Señores @Superdad y @JohnSwenson

Simple question for you. Given the LPS-1.2 can now handle a larger voltage range, from 5-12V, will it sound better at one or other end of the range? I ask this because I wonder if lower voltages incur more stages of regulation?

Looking forward to receiving my units to find out for myself!

Performance should be the same at all output voltage settings. The same cascade of ultra-low-noise linear voltage regulators is used regardless of the setting. We vary the level of charge given to the ultracapacitor strings. So the LDO regs always receive just a volt or so more than what they are being asked to produce.

That's how we are even able to offer the LPS-1.2 with the wide 5-12V output range. If we had to somewhere dump an extra 7 watts (7V/1A) of heat some where (to do 5V versus 12V), we would be in trouble!

JohnSwenson · January 26, 2018

1 hour ago, sandyk said:

Is the amount of effective capacitance the same at +5V as it is at +12V ?

Not that it probably matters anyway, given the large capacitance value of the Ultracaps.

The capacitor strings are exactly the same for all voltages, there are 7 ultracaps in series, each at 10F so that is 10/7 F per string all the time.

The string gets charged to 5V more than the output voltage. The output current pulls charge out of the string which causes the voltage across the string to drop. I arbitrarily chose 3V as the maximum drop, so when the string drops to 3V less than the charge voltage, the control circuit does a bank switch and the other cap string is connected to the output and the one that just got down to 3V below the charge level is now charged.

This minimum voltage across the strings is set to 1V above the output voltage of the bank regulator, which is 1 volt above the output regulator. The output regulator is always fed exactly 1V above its output. It varies for the bank regulator from 4V to 1V above the output voltage. This seems like a lot more power being dissipated by the bank regulator, but remember that a particular bank is only delivering power to the output half the time.

For a concrete example take 12V output, the capacitor string is charged to 17V (12+5), when that bank is connected to the output, the voltage is allowed to drop to 14V (17-3) at which point the banks are swapped. The bank regulator is set to 13V (14-1). The lowest point of the discharge cycle is 1V above the bank regulators output voltage. The output regulator is set to 12V (13-1) and always has exactly 1V from input to output.

As the output voltage is changed all these voltage points change as well so the voltage drop at each point in the system stays the same no matter what the output voltage.

So there are three adjustable voltage regulators and a bunch of ADCs monitoring the voltages all over the place in the system. This becomes way more complex because the charging system, the two banks and the output are all isolated from each other, you have to use opto-couplers to get control signals and data back and forth between them. What seems like a fairly simple system turns out to be a nightmare of complex details to make work.

John S.

sandyk · January 26, 2018

Hi John

Thanks for the detailed answer. Do you ever get much sleep ?

The amount of effort going into these designs from just yourself is mind boggling !

Kind regards

Alex

photonblur · January 26, 2018

Would the LPS-1.2 be compatible with a squeezebox touch

used as both a dac and server?

Superdad · January 26, 2018

18 minutes ago, sandyk said:

Hi John

Thanks for the detailed answer. Do you ever get much sleep?

He hasn't in the last few weeks, but now--after all the last kinks have been worked out--he is managing to get more rest. But his mind never stops!

18 minutes ago, sandyk said:

The amount of effort going into these designs from just yourself is mind boggling !

The topology and sophistication of the UltraCap design is indeed mind-boggling (for a little 1.1A LPS). Often I have to stand on my head to understand each of the circuits as John explains them to me.

(Even though I think I am useless, John finds it very helpful to talk the thorniest issues through in detail. And sometimes I'll ask a question from a different perspective that gets him thinking along a new line--and then he solves the problem! I like when that happens.)

There is a great deal of code in the FPGA to handle all the ins and outs of the UltraCap's operation. It is amazing what he can fix and change just in software.

JohnSwenson · January 26, 2018

2 hours ago, photonblur said:

Would the LPS-1.2 be compatible with a squeezebox touch

used as both a dac and server?

Yes it works great with an SBT, it makes a significant increase sound quality.

I'm not sure it can handle a USB powered hard drive plugged into the USB port though. I've never used my SBTs that way. If it is not powering a drive then yes it will work great with an SBT.

John S.

photonblur · January 26, 2018

2 hours ago, JohnSwenson said:

Yes it works great with an SBT, it makes a significant increase sound quality.

I'm not sure it can handle a USB powered hard drive plugged into the USB port though. I've never used my SBTs that way. If it is not powering a drive then yes it will work great with an SBT.

John S.

Great. If my info is correct the SBT has a 5.5mmx2.5mm dc barrel connector. Would this adapter

https://www.amazon.com/OMNIHIL-Adapter-Converter-Female-x2-5mm/dp/B01N812NKL/ref=sr_1_5?s=electronics&ie=UTF8&qid=1516947196&sr=1-5&keywords=5%2C5mm+female+to+2.5mm+male+adapter

be the one for the LPS-1.2 connector?

Thanks-

Superdad · January 26, 2018

1 hour ago, photonblur said:

Great. If my info is correct the SBT has a 5.5mmx2.5mm dc barrel connector. Would this adapter .....be the one for the LPS-1.2 connector?

Yes, that’s exactly the correct one.

tapatrick · January 26, 2018

On 1/21/2018 at 10:32 PM, Superdad said:

...I wish there was a way I could hear tour about and here some of what you folks have assembled.

Plus it would be wonderful to turn each other on to favorite musical artists...

Yes it would be amazing to be able to hear everyone's systems with so much experimenting and sharing going on it's a shame we can't do that.. or maybe someone is up for organizing a grand bus tour..

Superdad · February 2, 2018

Hello gang:

Just a quick update. My very efficient assistant has finished partially assembling 250 LPS1.2 cases (extrusion, front panel, and rubber feet), so now we are just waiting for the circuit boards to arrive.

John has been working hard fine tuning the software for the new boards (it is fabulously complex and makes my head spin when he describes the details of it to me). Unlike with the LPS-1, boot times for the LPS-1.2 will vary based on output setting (5V will complete charge and error checking quickly, 12V will take longer). Aside from the new features and improved performance, a goal of the LPS-1.2 was to make it more robust and reliable. Sure a 1% failure rate is not terrible (mostly from user charger over voltage)--15 out of 1500, but we would much prefer it to be around 0.4% or less. Repairs are a hassle for users and really costly for us--since we most always replaced the entire board. Based on the rigorous testing and error situation inducing that John has been doing, it looks like we will have achieved our goal. That makes me very happy!

Even the new, custom sized and printed product boxes are ready to go:

Since half our business is international, shipping costs and logistics are always on my mind. The US Postal Service provides us with poly/bubble Flat Rate Padded Pouches that, with insurance, we can ship anywhere in the world for $34. The same weight product in a non-Flat-Rate brown box would cost $65 via Priority Mail Intl. If we had to charge that much it would surely impact overseas sales.

Priority Mail is a reliable and tracked service, and we use a 3rd-party insurance carrier that pays quickly in the rare event of loss (good luck trying to get USPS to pay a claim on an international shipment). And in the USA, the same pouch via 2-day service anywhere in the country is just $9. (Of course we do offer FedEx and Express Mail/EMS service; and UPS in the USA--but it is rarely worth paying more for those services.)

The challenge is finding boxes that fit in these padded mailers and are large enough for the product. It is crazy, but all the stock mailer boxes are either too small or too large. I've spent countless hours researching boxes with wholesalers--to no avail. It is really silly considering all the firms that ship products. So I figured out the perfect size and had some custom made. Might change the color a bit on the next run, but I paid extra for the heavier 'B-flute' corrugated stock.

Thanks for a fantastic January everyone. We can't wait to get all these UltraCap LPS-1.2 units winging there way to you all!

We are still on track for shipping February 15th. The circuit boards are supposed to arrive on the 9th. It is going to be tight and require the prior weekend and several nights to do all the FPGA flashing (code load), testing, assembling, packaging, and my favorite (not) international shipping documentation.

The pre-order list is daunting--around 200 now--so it is possible that only the first 100 make it out on the 15th, with the rest to go each of the following days. I will keep everyone apprised.

Have a great weekend,

--Alex C.

agladstone · February 2, 2018

Alex:

Anxiously awaiting:) - (although if extra time is needed to ensure that it is 100% tested, etc , I’d prefer to wait an extra few days then get a unit that was “rushed” so to speak).

My system is just not the same without an LPS-1 in it! (I sold mine in anticipation of getting the new LPS-1)

Keep up the good work!

left channel · February 2, 2018

Another vote for letting the shipping schedule slip if helps avoid rushing any aspect of the testing. I'm in the first 80, but I wouldn't want to be one of the 0.4%.

Matias · February 2, 2018

Specially with international shipping and customs, letting a few days slip in order to improve reliability is a very desirable trade off. +1!

Superdad · February 2, 2018

8 hours ago, left channel said:

Another vote for letting the shipping schedule slip if helps avoid rushing any aspect of the testing. I'm in the first 80, but I wouldn't want to be one of the 0.4%.

Don’t worry folks: EVERY unit will be fully tested prior to shIpment.

Besides, all of what I spoke about (the testing and design changes to make the new model work perfectly) has already been done by John over the past months. That is all baked into the boards that are arriving. So assuming that the changes he made (parts value tweaks, some circuit logic adjustments) all made it perfectly into the board files and BoM (bill of materials), the boards will work beautifully. We both spent many hours triple-checking to be sure.

I won’t kid you and say I’m not a little nervous. There is after all a massive amount of our money and customer money riding on this. But this is an all or nothing thing: If one board works as expected, then every one of them should test fine. If a mistake was somehow made, then none of them will be right and the whole thing will have to be postponed (and I’ll be vomiting and wanting to hide in a cave ).

We’ll all know soon enough. But really, I am not too worried...

Chris_B · February 2, 2018

12 minutes ago, Superdad said:

Don’t worry folks: EVERY unit will be fully tested prior to shIpment.

Besides, all of what I spoke about (the testing and design changes to make the new model work perfectly) has already been done by John over the past months. That is all baked into the boards that are arriving. So assuming that the changes he made (parts value tweaks, some circuit logic adjustments) all made it perfectly into the board files and BoM (bill of materials), the boards will work beautifully. We both spent many hours triple-checking to be sure.

I won’t kid you and say I’m not a little nervous. There is after all a massive amount of our money and customer money riding on this. But this is an all or nothing thing: If one board works as expected, then every one of them should test fine. If a mistake was somehow made, then none of them will be right and the whole thing will have to be postponed (and I’ll be vomiting and wanting to hide in a cave ).

We’ll all know soon enough. But really, I am not too worried...

kilroy · February 3, 2018

Would be nice if this technology resulted in a full sized higher power LPS, say, a JS -2 replacement.

R1200CL · February 3, 2018

17 minutes ago, kilroy said:

Would be nice if this technology resulted in a full sized higher power LPS, say, a JS -2 replacement.

It’s only a question of you and others are willing to pay a lot....

It has been confirmed long time ago that it can “easily” be done.

Superdad · February 3, 2018

43 minutes ago, kilroy said:

Would be nice if this technology resulted in a full sized higher power LPS, say, a JS -2 replacement.

26 minutes ago, R1200CL said:

It’s only a question of you and others are willing to pay a lot...

Yup. A LOT. To get up into the 5-7 amp range--with voltages to 12V and above--you are looking at $3K+, in a case twice the size of a JS-2.

26 minutes ago, R1200CL said:

It has been confirmed long time ago that it can “easily” be done.

Glad you put that "easily" in quotes.

Announcing the next generation: UltraCap LPS-1.2 (Now 5/7/9/12V! ; Now with parallel LT3045 regulators!)

Recommended Posts

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Create an account or sign in to comment

Create an account

Sign in