Arcam Alpha 5 CDP upgrades

mjp200581

mjp200581

Joined
Feb 14, 2011
Messages
403
Reaction score
3
I'm currently having fun modifying an Arcam Alpha 5 CD player.

So far I have:

Replaced all of the electrolytics on the main PCB
Replaced all of the rectifier diodes
Replaced the broken tray loading cog
Replaced all of the electrolytics on the audio board
Replaced all of the op-amps with discete op-amps
Fitted the net audio super non-oversampling converter

If anyone is interested it would be great to start a thread on this player.

Has anyone here previously modified one of these players? Or are considering modifying theirs?
 
OK Dev,

You asked.....

I decided to buy an Arcam Alpha 5 after convincing myself that I wanted to try a TDA1541 based player. I wanted something which wasn't too expensive to buy (an therefore scary to work on) and also a player which had plenty of scope for DIY friendly modifications.

After reading around it seemed that a non-oversampling conversion was a key part of getting the most out of a TDA1541 player. Some 'NOS' conversions look very complicated but I knew that Net Audio made a drop in NOS converter board which simply replaced the SAA7220 filter chip. How simple can you get?

So a short-list was made of suitable TDA1541 which also have the SAA7220 chip and the Alpha 5 was my pick of the bunch.

I struck lucky when I was making a product enquiry with Net Audio about their NOS converters, they had a Alpha 5 for sale for £80 which already had the SAA7220 chip removed and a socket fitted ready to accept the NOS converter. Perfect!They even let me borrow one of their basic NOS converters for a few days to 'get me going'. Great service!!

When the player arrived the tray loading cog had died in transit so the first job was to replace it. Apparently this is a very common problem as the wrong type of grease was used in the factory which causes the plastic cog to degrade. A new cog and drive-belt were duly ordered from Hong Kong over ebay which arrived remarkably quickly. To replace the cog you need to remove the audio board and main circuit board. When you turn the main board upside down you can see two small torx screws which hold the pivot shaft in place. Remove these and the cog and belt can be replaced. I used a smear of silicon grease on the shaft only.



http://postimage.org/image/8x781apsv/

Initial listening impressions however weren't great. Compared to my other reference CD front end (see CA Discmagic/Isomagic thread) the sound was clearer with more detail but also sounded very harsh and edgy. The presentation had loads of power and attack but I found it very fatiguing and I found myself turning the volume down most of the time rather than up. In comparison the CA set up was much smoother, sweeter and generally more pleasant to listen to.

Here is a view of the standard player with the NOS converter fitted.

http://postimage.org/image/7gvpj5mvz/


There was however something promising about the sound so I decided to 'press on' and I'm glad I did.

More to follow
 
Part 2

So despite the unfavourable first listening impressions I decided to push on with the Alpha 5.

The borrowed NOS converter went back to Net Audio and I ordered their full SNOS converter upgrade kit.

This upgrade is restricted to the audio board only and contains:
14?x 220uF Rubycon ZLH (small black electrolytics)
1x Wima 10nF
4x Burr Brown OPA604 op-amps and sockets
1x Super NOS converter board (superior version, upgrade over standard NOS converter)
4x 35v 1000uF Rubycon ZLH
2x Elna Starget 100uF 35v
Instructions and some wire.

Here are the instructions:

Remove: SAA7220, Q2, Q102, LK201, R212, X20, C201, C202, R222, R251, C203, R224

Refit R224 to the underside of the PCB in the same position.

NB: if you fit the SNOS converter into a socket you don't need to remove R224)

Remove C224 and replace with the Wima

Connect the power pin of the SNOS to the cathode of D208

Replace Z102, Z2, Z101 and Z1 with the OPA604's fitted to sockets.

Replace C207, C208, C209, C213, C217, C226, C218, C210, C214, C228, C219, C221, C225, C108 and C8 with the 220uF Rubycons

Replace C205, C206, C241 and C242 with the larger 1000uF Rubycons

Replace C13 and C113 with the Elna Stargets. The +ve lead needs to be soldered to the track which leads to pin 6 of Z102/Z2 (cathode faces the transformer, anode to the transport mech).

Here is a pic of the kit fitted:


http://postimage.org/image/64e0h9pgf/

So how did it sound? Well in short it sounded crisper and more detailed than before but still sounded harsh and fatiguing. Somehow it didn't sound quite right.

Don't despair though because all of that changed for the better!
 
Part 3 The main board.

On most old CD players if the original electrolytic capacitors are still fitted they are likely to be passed their best and the player is likely yo benefit from having replacements fitted. The AA5 is no exception and it is quite common for some of the bigger electrolytics on the main board to fail.

I therefore decided to refurbish the main board of the AA5. Initially I was only going to replace the bigger power supply caps but in the end I decided to do the job properly whilst I had it apart and replaced all but one of the electrolytics and also most of the rectifier diodes. The only electrolytic I didn't replace is hidden under the transport mech which I couldn't be bothered to remove.

Here is a pic of the standard player with the audio board removed.

http://postimage.org/image/4duzfs7xb/

Here are details of the parts used:

C407, C406, C421, C408, C404 replaced with Rubycon ZLH 1000uF 35v

I hadn't had a previous experience with these capacitors but I figured that if Net Audio considers them good enough for their kits they must be OK. They also happen to be almost the same dimensions as the standard caps which makes replacement easy.

C146 replaced with Elna cerafine 470 uF 25v (I happened to have a spare)

C412, C604 replaced with 1uF 50v Panasonic FC

C409, C608, C411, C611, C605, C618, C619, C543, C544, C621, C620, C524, C530 replaced with 100uF 35v Rubycon ZLH

D404, D406, D405, D407, D402, D403, D411, D409, D410, D408, D418, D419, D401 replaced with Schottky 11DQ10

So after going to all that effort I plugged the player in with baited breath........

And it sounded fantastic! The fatiguing presentation had completely gone and the player sounded superbly toe-tapping and musical, full of 'boogie'.

I now realise that the Net Audio SNOS kit is superb. I suspect that the potential of the upgrade kit was being masked by some old and dodgy caps on the main board which were ruining the sound.
 
Part 4, Op-amps

I listening the the modified player for a few weeks to let the new components 'burn in' and to give the OPA604's time to settle down.

The player sounded great but I realised that I had accidentally ordered the kit with OPA604's rather than the more expensive option with supposedly superior 627's

If the player sounded this good with 604's how good would it sound with 627's?!

A quick search online revealed that OPA627's are around £20 each!

If I was going to spend that much I could 'bite the bullet' and instead of using 627's I could fit the discrete op-amps made by Dexa technologies (available from HiFi Collective). These are around £30 each so I chickened out and just ordered two which I fitted nearest to the RCA outputs.

The Dexa discrete op-amps only just fit under the lid of the player and I really mean ONLY JUST! You need need to be very careful when removed and replacing the lid.

They also overlap slightly when fitted so a used a strip of PVC electrical tape to make sure that they wouldn't short out between each other.

The discrete op-amps were a clear improvement over the 604's and improved noticeably over the next week or so. The sound of the discretes is very 'open' and superbly natural sounding. They have a really effortless sound which is not at all fatiguing. They aren't warm just very neutral and natural sounding.

They perhaps don't have the same sort of fast attack as the 604's and maybe sound less Hi Fi to some people but I for one much prefer them, so much so I ordered another pair!

While I was at it I also ordered 2x Elna Silmic II 100uF 16v capacitors to replace the Starget near the output which perform a DC blocking role. As far as I understand it Stargets aren't anything particularly special. The Silmic II is the current top of the range electrolytic from Elna and is considered by many to be one of the best available and comparable with Black Gates.

I only fitted the second pair of discrete op-amps and the Silmics yesterday so they are still settling in but initial impressions are good.

Here is a pic of the player as it stands.



http://postimage.org/image/uttr1o2zj/

Note that I have fitted a heat sink on top of the TDA1541A as it gets very warm in this player. I have also added two small silver coloured clip-on heat sinks to the previously 'naked' voltage regs as these too get hot.

Here is a close up.

http://postimage.org/image/xck1g3e3j/
 
Part 5, Buzz off!

Other Arcam Alpha 5 owners may have noticed that the transformer can be a bit noisy.

After a bit of investigation I decided that the top of the transformer was touching the underside of the lid causing the lid to buzz audibly. I therefore decided to remove the rubber block situated on the top of the transformer to create more clearance. I replaced the rubber block with a small self adhesive felt pad.

I also lined some of the case with sticky bitumen panels (actually it's roof flashing tape) in an effort to dampen any vibration in the metal casework. Be careful not to put this over the discrete op-amps as there isn't enough room!
 
Further upgrades?

I'm considering what further upgrades I might be able to make to the Alpha 5.

Any advice here would be welcome!

Here are some ideas:

1) Add a separate dedicated power supply for the SNOS converter which contains the clock crystal. The problem here is that the Alpha 5 doesn't have a regular on/off switch but instead has a standby mode. If a tap into the mains by the IEC power cord inlet the clock PSU would be on all of the time. Do you think this is an issue? Is there a better way to do this?

2) I'm a big fan of the fancy high bandwidth voltage regs available from Fidelity Audio, Dexa technologies etc. Has anyone used these in an Alpha 5? If so in which positions?

3) I'm thinking of swapping some of the resistors on the audio board for tantalum resistors (either Shinkoh or Audionote). Which resistors would be worth replacing? R13, R113?

Many thanks, Mike
 
Resistor upgrades?

Hi everyone,

Since the last post on this project I've made yet more changes to my Alpha 5 CDP!

I'll post details about these as soon as I've done enough listening to get a decent impression of the changes.

I'd like to experiment with some resistor upgrades in and around the audio output but I'm struggling to determine which components would be worth changing.

Here is a close up view of the area of the schematic which I'm concerned with.

Which resistors make be worth swapping out?




upload pics
 
Better regs than LM317T/LM337T?

I'd really like to try to improve the voltage regulators used on the audio board power supply.

There are four regulators providing +/- 5V and +/-15V these are LM317T and LM337T respectively.

Are there superior regulators which can be used as upgrades over the LM317T/LM337T?

As far as I can ascertain LT317A and LM317AT are closer tolerance alternatives to LM317T. Might they sound better?

Some people also seem to rate LT1085.

Here is the schematic:


pic upload
 
I wouldn't change any resistors for other types, they really have no sonic signature.

You could improve the 317 and 337 regulators by swapping out the voltage setting resistors ands replacing them with zener diodes. Have a read of this from about half way down the page "knowing this we can achieve even better..."

http://www.acoustica.org.uk/t/3pin_reg_notes1.html
 
Thanks Tenson,

That's a great article.

I might splash the cash on some discrete 'Super Regulators' instead. Like these:

http://www.newclassd.com/index.php?page=70

But all of the 'super' type discrete regs I have seen are all designed to be be drop-in replacements for 78xx type regs.

I think I could still use them with a few simple modifications to the circuit:

1) Remove the resistors highlighted in pink
2) Bypass (or replace with wire links) the resistors highlighted in green.


photo sharing

Could I then use something like the Dexa discrete regs? Oh and I'd need to be careful observe the correct pin configurations when making the swap.
 
If you go for those discrete regs, be sure they can supply enough current; they only supply about half that of a 78xx.

Personally I'd do as I said above though. You don't have lots of wideband noise on the PSU since it is not switch-mode. Simply lowering the output Z of the existing regulators with zener diodes will give you most of the benefits available from a better PSU.
 
Thanks Tenson,

I've been doing some background reading and I've got my head around how to set the output voltage of an LM317 reg using the formula:

Vout = 1.25 (1+ R2/R1)

In my case the standard components values are: (as shown in Fig 1)


image sharing

1.25v (1+ 2400/220) = 14.875v (near enough the 15V we're aiming for)

OK, so the benefit of using the zener diode (or an LED) is that they can provide a very accurate voltage drop with little impedance?

So the circuit would now look like Fig 2.

From what I can gather if using a zener diode in place of R2

Vout = 1.25 + Vz where Vz is the diode voltage (let's say 3.9v)

So 3.9+1.25 would give me an output voltage of 5.1v

Or with a 13.65v zener I'd have 1.25 + 13.65 = 14.9v

Is it really that simple?

But what about R1? How does that effect things and how do I calculate what the value of R1 needs to be?

I'm sorry if these questions are very basic, I'm on a steep learning curve!
 
Yes, you've got it :)

For R1, it should be as low as possible without taking too much power from the regulator or blowing up the diode. R1=100R is a good bet and that will push about 2watts through the diode if the reg output is set at 15V. This will take about 0.2amp from the regulator so plenty of the 1amp capability left for the CD player. Check the regulator temp and perhaps give it a clip-on heatsink if it gets hot.

The advantage of this method is low noise and low impedance. The voltage drop is not any more accurate.
 
OK the plan is starting to come together!

I like the idea using a normal diode in series with a zener to benefit from the temperature compensation effect. Assuming a normal diode has a voltage drop of around 0.6v I'll need a zener with a Vz of around 13v:

0.6+13+1.25=14.85

When I purchase the parts from farnell I'll check the data sheet specs and select the best components to get the value as close to 15v as possible.

For the +/- 5v supplies I like the idea of using two green LEDs in series each with a voltage drop of approx 1.94v

(1.94 x2)+1.25 = 5.13v

Tenson, would you be kind enough to have a look at the circuit diagram attached?

What values do you suggest for the 5v supplies R3 and R4?

Please assume that it was drawn by a joker with no knowledge of electronics (because it was) and point out any stupid mistakes which I may have made (diodes in wrong polarity etc).

Thanks again, Mike


images
 
I believe your diodes are the wrong polarity. Cathode to the adjust leg on the 317. Cathode to ground on the 337. R3 and R4 should also be okay at 100R I think.

I just noticed the 5V reg runs in series with the output of the 15V reg. I hpe the 15V regs don't get too hot.

I've not heard of using a normal diode along with a zener in this application. I wouldn't worry about temp stability, the absolute voltage is not of great importance provided it's within say 0.5V. If this concerns you, try a voltage reference IC instead of the zener (like LM431) but these will need setting resistors themselves.
 
Thanks again Tenson.

The idea of using a normal diode for temperature compensation came from the article you recommended to me in an earlier post:

www.acoustica.org.uk/t/3pin_reg_notes1.html

Specifically this bit:

One tip - using a zener in series with a normal diode (cathode to 0v), and you get nearly-perfect temperature compensation for free, too, because the zener avalanche voltage has a positive temperature coefficient, whilst the junction diode has a negative temp coefficient that nearly equates. The extra diode adds 0.6v to the output voltage of course.
Click to expand...

I'll be ordering the parts over the weekend and I'll report back when the job is completed. Wish me luck.
 
You might want to check if both diodes or only the standard one need cathode to 0V. Felix of this forum wrote that article so he's the man to ask.
 
Hi Tenson,

I did PM Felix and he was kind enough to send me this reply:

No problem - yes, you do want the anode (the 'bar' on the symbol) of the zener pointing 'up' in your diagram in both positions: so that's anode to Vadj pin of the LM317 and anode to 0v for the LM337.

That's because the zeners work in 'breakdown' /avalanche mode. If you find you only have about ~3 v out of each reg,. then the zener is the wrong way round (when it will have the normal 0.6-0.7v 'forward' voltage). It *won't* do any harm, but you wont get any sound out of your cd player. But the regular diodes need to point in the opposite direction in both cases to conduct at all.

Note you won't get exactly 15.0 and 5.0v out as a result using these mods, but within 0.2 or even 0.3vdc is just fine; absolute dc voltage doesn't matter much.

R1 in each case can be anything convenient, 100-240R, you can leave what is fitted as standard if you like.

Note all these tricks do is make the adjustable reg perform as the ideal 1.25v reg the datasheets is written around. It can't do better than that.
Click to expand...

I really appreciate the kind help I'm getting with this.

Here is the (hopefully) corrected circuit diagram.


windows 7 screen shot
 

Latest posts

Back
Top