advertisement


The John Westlake/Lakewest MDAC/FDAC, VFET and Detox

sq225917 said:
Far easier to specify the hardware, flow chart the functionality and hand it off to a real coder.
Click to expand...

UI stuff yes, not the hardware control blocks, Ill basically create libraries for the hardware level stuff - isolating the more complex (hardware) side from the programmer. The idea is a programmer then just calls functions and transferred data if required, to and from the hardware.
 
John,

I have an offer of assistance from a very experienced commercial developer colleague, who has C++ among his talents. I've worked with him and he's very effective - I've seen him pick up complex novel problems in other people's code, identify the issue then resolve it, in minutes. If nothing else, he could advise on your efforts and potentially stop you before you go up a coding blind alley.

I believe this offer has reached you via another respected member of this forum but I'd like to repeat it here. We'd all be pleased to see you concentrate your efforts on the things where you can add value - learning C++ is not one of them so please give this some thought. Happy for you to PM me if you wish to explore further.

Trevor
 
Hi Trevor,

Thanks for the offer - as I explained to the "other respected member" :) what I need is a programmer with hardware experience who can handle the XMOS, the C++ is just for the MCU handling the UI / front panel, this is not of great concern as its easy to find skilled programmers in C++ / UI design. XMOS requires a person skilled in Hardware + software.

Jarek worked with me for over 4.5 years on the project so it is really very hard to find someone who can replace him in short order and has the skills ready to pickup from where he left off. The Company I'm now working with has some XMOS experience, but only at a more basic level - but its the best option I can find at the moment - so I'm simplifying the design to reduce the delay.

As I mentioned in my earlier post, to ease the requirements on the XMOS programmer - the UI can be partitioned off onto a separated front panel MCU - with a communication interface between XMOS and Front panel MCU. Its messy, and creates problems when updating software (two different software payloads, how to achieve this)...

Once we have partitioned the XMOS hardware / UI functions and created an API / interface between the MCU's then I'm more then happy to have 3rd party's work on the UI front end (in C++)... I initially plan a very basic UI to get the PCB's out as soon as possible - this can be updated with improvements over time by the user via software downloads.

I need to be clear, that without an in house software guy, then i need to learn the programming skills to power-up the hardware and bench test the designs... Once I have hardware with basic functionality proven / tested then the higher level software can be written. At a hardware level a "programmer" needs to understand and have the test equipment to debug the electronics, otimise PID control loops, work with VHDL (for the FPGA code), work with MathLab / Simulink to design the digital filters / modulators, understand the USB stack etc. this is NOT simply a task for your average C++ software programmer this is VERY much hardware oriented "Software"... and hence why without other options I'm tasked to complete...

The company I'm working with is saying 5-6 months - which seems far considering Jarek work with me on the design for over 4.5 years...
 
Last edited:
Hi John,

Thanks for your response.

Your reply gives me no less reason to think this offer should not be pursued, at least to the next stage (see below).

In these unprecedented times, a large proportion of us are working from home for the foreseeable future. The luxury and utility of another in-house programmer cannot be assumed again but we are where we are and I don't see that as a disadvantage because remote working is now entirely viable.

Commercial-standard programming differs from amateur programming because, amongst other things, hard lessons have been learned from experience; good established practices are incorporated, and efficient and effective solutions are produced. I would also observe that commercial programming practices don't risk scuppering the project when a key resource leaves.

I'm sincerely hope you would agree that I should no more let you loose on any of my company's coding requirements than our forum members would feel comfortable about you spending your valuable time continuing to pursue this direction unaided.

One of the most critical things to get right in any development is a clear and unambiguous specification of requirements. I expect that if you could provide this, and advise on any necessary constraints or boundaries, any decent programmer could advise on an appropriate approach/structure, etc., if not actually code it; crucially, they could also advise on how they definitely would NOT do it.

In non-IT speak, if you gave an experienced car driver any powered vehicle with a steering wheel we can probably all agree that they'd be a whole lot safer at driving the tractor/HGV/speedboat/forklift truck/combine harvester, etc. than someone who hasn't yet got a driving licence. Experience - you just can't beat it.

Any experienced programmer will tell you it takes a few years before they're really effective (if they ever make that level) - I'm offering you a proven good one. Pragmatically, your time is precious and you're far more likely to succeed with a helping hand. My suggested next step would be a 1-2-1 discussion with aforementioned resource with the aim of establishing viability or, potentially, not.

Still watching for incoming PM's...
 
As a non-contributor please excuse my posting on the thread. However, I’m a software developer with nearly 40 years experience, so just wanted to add my support for the guys who are cautioning against a novice person getting involved in actual coding of any sort.

It really is a case of not knowing what you don’t know.

Best wishes to all involved.
 
Paid to level 4 mdac2 investment, £400 I do believe .
Just to make whoever it may concern aware of the fact .
It has been a long wait ...
 
Probably working at the clock unit and gotten into a time loop.

Mdac will have been ready for delivery last year at christmas (futur tense perfect).

Wikipedia is almost right here:
Time loop

...
A time loop or temporal loop in fiction is a plot device whereby characters re-experience a span of time which is repeated, sometimes more than once, with some hope of breaking out of the cycle of repetition.

Well they don't seem to know that it exists not only in fiction, and hope of breaking out is very limited to say the least...
 
  • Like
Reactions: JTC
Hi,

Apologies for being off topic, but can anyone tell me which torx screws I need to unscrew from the back panel to take the lid off? I want to check if my M-Dac's caps have started to bulge.
 
Jeroengrt said: ↑
Is there any news on the mdac2? when would it be available?


I'm currently working on the PCB design... I cannot say how long - but first PCB's in say 60 days...

JohnW, Jun 9, 2018

so pcb design lost?, Jarek did a new one in a different cad tool and now John is back designing another one?, no backups? we were so close to completion....
 
NONE.
Four each side, none of the others. Ease front and back panels out gently just until they are free. There will then be enough slack in the connecting ribbon cable to tilt the front panel and slide it through the shell, without having to disconnect anything. Easier than it sounds.
 
russel said:
Jeroengrt said: ↑
Is there any news on the mdac2? when would it be available?


I'm currently working on the PCB design... I cannot say how long - but first PCB's in say 60 days...

JohnW, Jun 9, 2018

so pcb design lost?, Jarek did a new one in a different cad tool and now John is back designing another one?, no backups? we were so close to completion....
Click to expand...

John has provided some updates, but Jarek has left the project for unknown reasons. That leaves John with a gap in programming the Firmware for the new DAC and he's pretty much doing everything as he doesn't seem to want to accept help from people on this side of the fence looking over his shoulders. He is apparently currently redesigning the PCB layout for reasons like you we do not understand on that CAD layout. My own thoughts are that having again committed to use the MDAC2 chassis, he is space constrained and is trying to shoehorn it all in to fit the chassis. Personally, if it makes the design and project easier and faster, then I wish he would just use a bigger existing commercially available chassis. I have suggested this to him. Other than this, John is designing, learning to programme firmware, managing the supply chain, project management et al. He is looking for someone to help with the big Firmware programming, but they must be able to work to his design and requirements. The Firmware programming is estimated at 5-6 months and that's once a PCB design is finalised. At this rate, we will be lucky to get anything before 2022, but it might be very worth the wait if we are still alive when and if it does arrive. My advice to those holding out for an FDAC would be to get the MDAC2 first and then upgrade to an FDAC if it ever materialises.
 
Thank you. That is the most encouraging post for quite some time... I signed up for FDAC, but I do not possess an MDAC. If anyone happens to have a spare one to sell (as I seem to recall that MDAC2 is a basically a drop-in replacement board -is that right?), then please PM me... Otherwise I guess I will just have to hope the FDAC will someday appear...
 
KevinNZ said:
Thank you. That is the most encouraging post for quite some time... I signed up for FDAC, but I do not possess an MDAC. If anyone happens to have a spare one to sell (as I seem to recall that MDAC2 is a basically a drop-in replacement board -is that right?), then please PM me... Otherwise I guess I will just have to hope the FDAC will someday appear...
Click to expand...
No need for you to go out and buy a MDAC!
There will also be a MDAC2 chassis for the rest of us that no longer have the original MDAC.
Personally I prefer a new chassis.

FDAC will be considerable more expensive and be made after MDAC2 and VFET amps.
 
I've no longer got a chassis myself, so this is good news I think. And probably a new chassis would be a nice thing. I do still have my defunct (I think) TOY L2 mainboard, which needs re-capped, but no case or PSU. I won't be holding my breath, however, but if it does come to pass it might end up being a better DAC than my current one so possibly worth buying.
 
...and hello everybody. I'm just making my 2020 visit to check how's MDAC2 going.
It's checked. No good news. I guess with a lot of luck, we might see the first units by xmas.
(not saying which year, though)

I hope there no casualties amongst the waiting list. My regards to everybody!
Stay healthy.
 
You must log in or register to reply here.


advertisement


Back
Top